Home

User`s Manual - TTC Technologies

1. 29 represents one or more lines of scripting language segment iiis incremented by 1 and the body of the loop executed until expression is greater than expression2 Other loop and decision structures can be found in the iSCRIPT Language Reference in Appendix A The complete program for Problem 4 2 is shown in Figure 4 6 If you compare this code with Figure 4 5 you can understand the meaning of each part of the code The code can be found in the subfolder SampleScripts AircraftDrag Example4 2 of the iSCRIPT installation folder WOIHObwWNH AirplaneDrag isc Program main drag rho V as real Cd as real Vores 11 drag_res 11 as real i as integer drag 20000 rho 1 0e 6 kA 100 0 4470 A 7 0 ou Ca 2 drag rho V 2 A do i 1 11 V_res i 200 0 4470 10 i 1 drag_res i Cd rho V_res i 2 a 2 end do end program Figure 4 6 An iSCRIPT program for evaluating the drag on an aircraft at a range of low speeds To run this code you can carry out the following steps 1 Open iSCRIPT Editor by selecting the program from the Windows Start menu 2 Select File gt Open from the menu gt The File Open dialog box appears 3 Navigate to the subfolder SampleScripts AircraftDrag of the iSCRIPT installation folder 4 Open the file AircraftDrag isc Select Tools gt Run Current Script Project File from the menu 6 The results will be sho
2. 96 AddObjective component variable maxmin Note Segments enclosed in square brackets are optional and may be omitted component The component to be optimized a string limited to 24 characters This component must be a component previously declared with the CreateComponent command variable Name of the variable a string limited to 24 characters This variable must be a variable previously declared for this component using the CreateVariable command maxmin O or 1 Indicates whether this is a minimization or maximization objective Use 0 to minimize this variable and 1 to obtain a maximum This argument is optional The default value is 0 NOTE THERE IS NO NEED TO ENCLOSE CHARACTER ARGUMENTS IN QUOTES FOR THIS COMMAND Example CreateComponent Heat_Ex1 CreateVariable Heat_Ex1 Tin CreateVariable Heat_Ex1 Tout CreateVariable Heat_Ex1 Length CreateVariable Heat_Ex1 Width CreateVariable Heat_Ex1 Weight AddObjective Heat_Ex1 Weight 0 B 2 2 Indicating an Free Variable for Optimization A variable to be varied in the search for an optimum is indicated as follows AddVarObjective component variable delta Note Segments enclosed in square brackets are optional and may be omitted component The component to which the variable belongs a string limited to 24 characters This component must be the component to be optimized or a sub component of it This 97 component must be a component previ
3. unit A string representing the engineering unit for the variable e g m s This argument is optional and is should be limited to 20 characters if provided A component may have several variables A component variable can be referenced by using 37 ComponentName VariableName Example Cylinder radius 5 Cylinder height 10 Cylinder Area 2 pi Cylinder radius Cylinder height A component variable can be used and its value can be changed in any program subprogram or component Each component must have a unique subroutine that has the same name as the component name The component subroutine requires no argument This subroutine is automatically executed once the Execute command is called Component_Name Execute or Call Component_Name Execute The component may also be optimized by using the Optimize command as follows Component_Name Optimize or Call Component_Name Optimize With the above statements iSCRIPT provides a component modeling method for the design and optimization of a complicated system A large system can be decomposed into several components and each component is modeled in its own executable subprogram We will give an example in the next section to illustrate how to use this component modeling technique to solve a real problem 38 5 2 Creating Subsystems and Systems in iSCRIPT Conceptually subsystems have the same data structure and propertie
4. 70 Global maxInitialEvaluations 200 Global maxGenerations 10 134 Global optconvergencelimit 1 0E 7 Global mutationfreq 0 2 Global sampsizepervariable 90 Rastrigin Optimize end program subroutine Rastrigin This subroutine is an execute model for entire subsystem and system x1 x2 pi As Double xl Rastrigin x1 x2 Rastrigin x2 pi 4 atan 1 0 pi 3 1415926536 Rastrigin y 20 0 x1 x1 x2 x2 10 cos 2 pi x1 cos 2 pi x2 end subroutine Output The output file again is outputscript txt The correct results were obtained in about 2 9 seconds on a Pentium workstation using only a population of 70 realizations Details of the optimization process are recorded in the file optimize txt The details include the initial values of the optimization variables and the objective the various realizations of the system being evaluated the array of viable systems or realizations population of individuals in genetic algorithm parlance by generation or as the optimization progresses and the final results 135 References Munoz J R A Decomposition Strategy Based on Thermoeconomic Isolation Applied to the Optimal Synthesis Design and Operation of an Advanced Fighter Aircraft System M S Thesis Advisor M R von Spakovsky Mechanical Engineering Dept Virginia Polytechnic University Blacksburg VA February 2002 Alabi K Ladeinde F von Spakovsky M R Moorhouse D Camberos J As
5. Cdo_Sup 23 this code uses the component built method for supersonics guiven by Raymer 24 25 Sref 325 26 V 856 27 blogic 2 2 2 jj 1 28 ik blogic 2 4 2 jj 29 Swet_fuselage 588 31 fuselage 33 Wo 24000 34 45 0 3048 a Wo c 0 3048 35 k 0 052 10 5 for smooth composite 37 alt 1800 38 mu 0 00008 39 rho 1 25 40 M 0 85 41 42 R_fuselage rho V I mu 43 44 R cutoff sub 38 21 I k 1 053 eq 12 28 Raymer 45 R_ cutoff sup 44 62 I k 1 053 M 1 16 eq 12 29 Raymer 46 Cf fuselage 0 454 log10 R_fuselage 2 58 1 0 144 M 2 0 65 47 48 Amax 20 9 294 Sref 3 83 49 d 5 5 294 Sref sqrt 4 pi Amax 50 f fuselage 1 0 3048 d 51 FF_fuselage 1 60 f_fuselage 3 f_fuselage 400 53 Cdo_fuselage Cf_fuselage Swet_fuselage Sref FF_fuselage 54 55 end program 116 Output FINAL VALUES OF VARIABLES Logical Variables 4 blogic 3 2 FFFTFF brun F lexist F lrun F Integer2 Variables 3 il 0 j3 0 j4 0 Integer Variables 8 icount 0 ik 1 jj 0 kk 0 iunit 0 ifile 0 i 0 j 0 Real Variables 19 Sref 325 0000 Wo 24000 00 1 13 71600 k 5 2000001E 07 Swet_fuselage 588 0000 alt 1800 000 mu 7 9999998E 05 rho 1 250000 R_fuselage 1 8345150E 08 R_cutoff_sub 2 4930214E 09 Cf_fuselage 1 8315958E 03 Amax 19 27374 d 6 079932 f fuselage 7 401399 R_cuto
6. Se Q A A A a A WO e 14 15 15 16 19 22 22 23 26 31 35 35 38 38 40 41 43 Table of Contents cont d 6 Optimization in iSCRIPT 6 1 6 2 6 3 Optimization Based on Component Modeling Optimization of a System with a Single Component Optimization of a System with Multiple Components 7 Running iSCRIPT in Parallel 7 1 T2 7 3 System Requirements Running an iSCRIPT Program in Parallel A Sample Optimization Problem Running in Parallel 8 Interfacing with Other Software 8 1 8 2 8 3 8 4 8 5 8 6 Purpose Running a Third Party Software or Executable Setting Changing the Working Directory Determining the Working Directory Example of Running a MATLAB Script Running an Executable or Third Party Software 9 Conclusions Appendix A iSCRIPT Language Reference A l A 2 A 3 A 4 A S A 6 A 7 A 8 A9 Variables and Expressions Arrays Decision Structure Loop Structure Subprogram and Function Other Program Flow Structure Intrinsic Functions Input Output Component Modeling iii 49 49 51 54 60 60 60 61 64 64 64 65 65 65 68 71 72 72 76 77 77 79 83 85 85 89 Table of Contents cont d Appendix B iSCRIPT Optimization Reference B 1 Design Optimization Analysis Procedures B 2 Procedures for Performing a Detailed Optimization in 1iSCRIPT Appendix C More Sample Problems References iv 93 93 94 1
7. The system is illustrated in Figure 5 8 fc Condenser Cooling Tower ac A m Figure 5 8 Cooling tower pump and condenser system of a heat rejection system The objective is to evaluate the initial plus operating costs of the system The heat rejection rate from the condenser is provided as 14MW The following costs in dollars are included in the problem description e Initial cost of cooling tower 800A where A area m e Lifetime pumping cost 0 0005 w where w flow rate of water kg s e Lifetime penalty in power production due to elevation of temperature in cooling water 270t where t temperature of water entering the condenser C e The rate of heat transfer from the cooling tower can be represented adequately by the expression q 3 7w tA W Assume A 170 m and w 200 kg s What is the total cost of the system 44 The heat rejection system can be physically decomposed into three components cooling tower pump and condenser as illustrated in Figure 5 8 Note that in a detailed model each of the components may be subsystems comprised of other components Heat Rejection System Cooling Tower Pump Condenser Component Component Component Figure 5 9 Decomposition of the system By virtue of the physical decomposition strategy adopted the total cost Csys can be represented as Ci Cro PC pu E Coning 5 1 For the cooling tower the cost is Cone 8004A 5 2 For
8. e The output of the changedirectory command is 1 if successful and 0 if an error occurred 8 4 Determining the Working Directory To determine the current directory at isCRIPT runtime iSCRIPT has provided the keyword currentdirectory When used with the write statements Section 4 8 the current working directory is printed See sample problem 18 which is described in the appendix 8 5 Example of Running a MATLAB script A short MATLAB script that uses some of the matrix manipulation commands in MATLAB is used to the extent to which iSCRIPT can execute MATLAB programs Model 423 211 410 x 4 6 6 y 1 2 1 b 2 1 2 PSO JE LZ ee a a Z yxx bxx 2y x 67 8 5 1 Running the MATLAB script The MATLAB script is shown below z 2 y x a x 2 sin y Note that Lines 4 and 6 have the ending removed so that MATLAB would print an output to screen The file containing this script is matlaba70 m in the subfolder SampleScripts MATLABprograms of the iSCRIPT installation folder There are two ways to run this script in MATLAB 1 Method 1 e Open the MATLAB program e Simply copy and paste the above script or from the open matlaba70 m file into the MATLAB command window e The output values of the matrix Zand are printed out 2 Method 2 e Open the MATLAB program e Set the Current Directory to the SampleScripts MATLABprograms subfolder of the iSCRIPT installation folder as shown in Figure 7 1
9. rho 1 05 v 17 0e 4 L 5 mu 8 5E 5 Models Re Reynolds rho v L mu mu 8 5E 5 call Computef Re f if Re then testing a logical statement if Re then testing a logical statement if L 1 then 0 001 elseif L 2 then 0 002 elseif L 3 then 0 003 elseif L 4 then 0 004 elseif L 5 then 0 06 Re 0 4 elseif L 6 then 0 006 else 0 007 end i end end L 4 ResetL i 1 while i lt 2 L then f f 0 01 Re Re 0 01 i i 1 testing a while loop segment end Comment line here L 5 end program function Reynolds rho u L mu rho u L mu Reynolds as real Reynolds rho u L mu u 12 2 end function subroutine Computef Re f Re f as real if Re lt 2500 then f 64 0 Re Testing an if segment in a subroutine end if Re gt 2500 then Testing a nested if segment in a subroutine f 16 0 Re 0 4 end end subroutine 111 Output Logical Variables 0 Integer2 Variables 0 Integer Variables 0 Real Variables 7 rho 1 050000 v 12 20000 1 5 000000 mu 8 5000000E 05 re 105 0700 f 7 9325572E 02 i 8 000000 Double Variables 0 112 Sample Problem 6 This script is used to illustrate the use of while statements and elseif and else statements within subroutines and functions Model _ PUL u Re aan if Re lt 2500 Re f 16Re if Re gt 2500 if Re f Re
10. 4 2 3 4 6 6 7 8 9 y 2 1 1 1 2 1 1 1 2 b 410 212 51 1 Zz y x b x 2 y x z 2 y x a x 2 sin y J J The file containing this script is matlaba70 isc in the subfolder SampleScripts MATLABprograms of the iSCRIPT installation folder Run this file in iSCRIPT as follows 69 1 Open iScript Editor 2 Select File gt Open from the menu gt The File Open dialog box appears 3 Navigate to the subfolder SampleScripts MATLABprograms of the iSCRIPT installation folder 4 Open the file matlaba70 isc 5 Select Tools gt Run Current Script Project File from the menu The program runs and the output values of the matrix z are printed out 8 5 3 Running other MATLAB sample programs Additional examples running scripts created in MATLAB are available in the subfolder SampleScripts MATLABprograms of the iSCRIPT installation folder and are described in sample problems 12 through 15 in Appendix C 8 6 Running an Executable or Third Party Software The executable used for this illustration was actually generated from a MATLAB script The procedure to compile the script into an executable is also described The method for interfacing with a third party software is illustrated in a system modeling context in which one or all of the subsystems are modeled in a different software Figure 7 2 presents a schematic of the procedure In the figure a complete aircraft is modeled m
11. 8 5E 5 call Computef Re f if Re then testing a logical statement if Re then testing a logical statement if L 5 then f 0 06 Re 0 4 end if if L 4 then f 0 06 Re 0 4 end if end end L 4 ResetL for i 1 2 L testing a loop segment f f 0 01 Re Re 0 01 end Comment line here L 5 end program function Reynolds rho u L mu rho u L mu Reynolds as real Reynolds rho u L mu u 12 2 end function subroutine Computef Re f Re f as real if Re lt 2500 then f 64 0 Re Testing an if segment in a subroutine end if Re gt 2500 then Testing a nested if segment in a subroutine f 16 0 Re 0 4 end end subroutine 109 Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 0 Real Variables 7 rho 1 050000 v 12 20000 1 5 000000 mu 8 5000000E 05 re 105 0800 f 8 9325570E 02 i 9 000000 Double Variables 0 110 Sample Problem 5 This script is used to illustrate the use of while statements and elseif and else statements Model Re wus u f a if Re lt 2500 Re f l16Re if Re 22500 if Re fC Re f 0 001 L 1 f 0 002 ES2 f 0 003 L 3 f 0 004 L 4 f 0 06Re L 5 f 0 006 L 6 f 0 007 any other value of L i l f f 0 01 while i lt 2L Re Re 0 01 i 1 Input File eqguations6f isc program rho v L mu Re f i as real Initial Values
12. Line 4 is the statement used to create the component and Lines 5 through 8 describe the variables of the component Note that here we have used only three parameters with the Create Variable statment component name variable name and variable type We chose to not supply other additional optional properties of the variables 40 including providing a default value setting limits or constraints on the variables or providing an engineering unit The default value in this case will be automatically set to be zero In later problems we will demonstrate how to use those properties of the iSCRIPT component variable CircularCylinder isc Program main 1 2 3 Create the Cylinder component and its variables 4 CreateComponent Cylinder the circular_cylinder S5 CreateVariable Cylinder radius double 6 CreateVariable Cylinder height double 7 CreateVariable Cylinder surface_area double 6 CreateVariable Cylinder volume double Cylinder radius 5 Cylinder height 10 Cylinder Execute end program subroutine Cylinder i pi as real pi 3 141592654 Cylinder surface_ area 2 pi Cylinder radius 2 2 pi Cylinder radius Cylinder height Cylinder volume pi Cylinder radius 2 Cylinder height end subroutine Figure 5 2 iSCRIPT program for Problem 5 1 Line 13 contains a statement to evaluate the component Lines 17 through 26 represent the actual model of the component implemented as a subroutine with
13. The entire system may be modeled by modeling calling each component model each of its components The system model may also include interactions between the components of the system Figure 3 2 summarizes the basic procedures for performance analysis of an engineering system in iSCRIPT using a simple gas exhaust system for illustration The basic steps are also listed below 17 Step 1 Step 2 Step 3 Step 4 Step 5 Step 6 Step 7 Step 8 Open iSCRIPT Editor by selecting iSCRIPT Editor from the Windows Start menu Create a new file in iSCRIPT Editor and save it as an isc file Type the words program programname and end program to create a main program In the body of the main program a declare all the system s components and their variables To declare a component use the CreateComponent statement To declare a component variable use the CreateVariable statement b declare the overall system and its variable using CreateComponent and CreateVariable statements c assign the input values to the system or component variables by typing ComponentName VariableName expression d type SystemName Execute to evaluate the overall system After the main program is written type words subroutine ComponentName and end subroutine to create the component subprograms for each of the system components In the body of the subroutine a type all the equations mode
14. a 2 1 3 b y x xa b y x a Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 3 x 3 3 4 2 3 4 6 6 7 8 9 y 3 1 1 0 0 a 1 3 1 3 Real Variables 1 b 0 0000000E 00 Double Variables 0 File Edit Yiew Text Debug Breakpoints Web Window Help Dee i eeno SMF BO BS su pe matlaba70 m matlaba10 m Ln3 Col15 y 2 1 1 1 2 1 1l 1 2 AM b 410 212 511 z y x b x 2 ty x a x 2 sin y a 2 3in y H 8 22 06 7 51 AM gt Gita Mi 124 Sample Problem 14 MATLAB Compatibility This script was used to illustrate compatibility with MATLAB matrix manipulation features The equivalent MATLAB file is matlaba70 m Model Input File matlaba70 isc i x 3 3 y 3 3 b 3 3 as integer 2 3 3 a 3 3 as real x 4 2 3 4 6 6 7 8 9 y 2 1 1 1 2 1 1 1 2 T b 410 212 51 1 Z yxx bxx 2y x oo a x sin y 4 2 3 2 x 4 6 6 y 1 789 1 Se N e N e e e II An A e me N O Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 4 i 0 x 3 3 4 2 3 4 6 6 7 8 9 y 3 3 2 1 1 1 2 1 1 1 2 b 3 3 4 1 0 2 1 2 5 1 1 Real Variables 2 z 3 3 41 66667 35 00000 36 33333 47 66667 54 00000 49 33333 50 33333 43 00000 61 66667 a 3 3 5 818595 3 682942 4 682942 5 682942 7 818595 7 682942 8 682942 9 682942 10
15. component_name Execute statement Type the interaction equations between components Type the overall system performance equations Run the program by selecting Tools gt Run Current Script Project File program main Declare N components of system by using CreateComponent Declare the component variables for each component by using CreateVariable Declare the overall system by using CreateComponent Declare the overall system variables by using CreateVariable Assign the input values for the system or component variables Execute the overall system by using System Execute program subroutine componenti Type all the equations for the component 1 end subroutine subroutine componentN Type all the equations for the component N end subroutine subroutine system Execute component 1 by Componenti Execute Type the interaction equations between component 1 and component 2 Execute component 2 by Component2 Execute Type the interaction equations between component N 1 and component N Execute component N by ComponentN Execute Type the overall system performance equations subroutine Example A gas exhaust system which consists of two components pipe and pump The costs of pipe and pump are 160D and 2 2 x 108w respectively where D is the diameter of pipe and w is the gas mass flow rate of pump Evaluate the overall system cost Assume D 110 and w 0 020 program main Create p
16. friction 3 794308979507086E 002 rougha 1 000000000000000E 004 eps 1 000000000000000E 004 rselect 1 000000000000000E 004 doveps 100 000000000000 epsovd 1 000000000000000E 002 118 facl 5 14000000000000 fac2 1285 22349993048 fac3 1 00723609551218 fac4 6 262561683476402E 003 fac5 5 13373743831652 fac6 0 194789860606491 fac7 112 286369433646 fac8 135 300985397874 fac9 15192 4564311220 deltaf 2 070842958548824E 011 depst 1 000000000000000E 004 g 3 794308979509621E 002 r 2 073378274625171E 011 small2 1 000000000000000E 009 forlog 10 0000000000000 fsuggest 3 794308979507086E 002 drdf 1 00122429181116 flowleft 4 000000000000000E 004 flowmid 2 000000000000000E 004 flowmid2 5 000000000000000E 005 enorm 2 978714945569056E 019 Double Variables 0 FORTRAN program Results from FORTRAN execution of moodytest for friction 3 794308979507086E 002 fric 3 794308979507086E 002 fric2 3 804064733943083E 002 fric3 9 699909063352531E 005 ilam 2 rey 659800 000000000 119 Sample Problem 9 This script was used to illustrate the use of the natural recursive characteristics of iSCRIPT functions The script is used to calculate the factorial of a number The factorial of 4 was computed Model Input ni n n 1 2 1 File equations6kk isc Ol 1 Recursice function test using the factorial of a number program mumber rfactorial as integer Initia
17. 0 localemCp global emCp end program 94 Appendix B iSCRIPT Optimization Reference B 1 Design Optimization Analysis Procedures iSCRIPT provides functions and procedures to optimize components and systems The functions utilize a combination of genetic and gradient based algorithms The integrated local global optimization ILGO procedure is a powerful option used to optimize a system consisting of several sub systems This procedure allows the optimization of large systems within a feasible time frame as compared to procedures that utilize nested optimization loops through several component optimization levels The procedure is illustrated below Coupling between subsystem 1 amp L A ILGO Coupling between Subsystem 1 gt Subsystem 2 Coupling between subsystem 1 amp 2 subsystem 2 amp L gt Subsystem L A DETAILED Component 1 Component 2 Component n Component n 1 Component n m Component n m k 1 Component n m k 2 OPTIMIZATION OPTIMIZATION f Figure 5 1 Optimization procedure DETAILED OPTIMIZATION When an optimization command is invoked on any component the system will launch a detailed optimization based on a combination of genetic and gradient based algorithms The relationship between sub systems systems and components is utiliz
18. The procedures required to solve this problem follow the general procedure presented in Section 3 3 The steps are as follows Step 1 Open iSCRIPT Editor Step 2 Create a new file and save it as an isc file Step 3 Type the words program main and end program to create a main program In the body of the main program a declare the Cylinder system by using CreateComponent statement b declare the four component variables radius height surface_area volume of the Cylinder system using CreateVariable statement c assign the input value for Cylinder volume e declare the objective variable Cylinder surface_area by using AddObjective statement f declare the optimization variable Cylinder radius by using AddVarObjective statement g type Cylinder Optimize to optimize the overall system 53 Step 4 After the main program is written type the words subroutine Cylinder and end subroutine to create the system subroutines for the Cylinder component In the body of the subroutine a type the equations for the height and surface_area Step 5 Save the file Step 6 Select Tools gt Run Current Script Project File from the menu to run the iSCRIPT code Step 7 View the output on screen or in the file outputscript txt Note that the only input variable for this problem is the volume The task is to determine values of the other variables r and A that minimize the surface area of the mode
19. UL 1 rho v L mu Re f i as real Re eae 2 u 3 Initial Values 4 64 5 rho 1 05 6 v 17 0e 4 f if Re lt 2500 7 L 5 2 Re 7 8 mu 85E5 f 16Re if Re 2500 9 Hi x 10 Models f 16Re if ReANDL 5 1 12 Re rho v L mu 13 f f 0 01 14 loop i 1to2L 15 if Re lt 2500 then Testing an if segment Re Re 0 01 16 f 64 0 Re 17 end 18 if Re gt 2500 then Testing a nested if segment 19 f 16 0 Re 0 4 20 end 21 if Re then testing a logical statement 22 if L 5 then 23 f 0 06 Re 0 4 24 end if 25 end 26 27 L 4 ResetL 28 29 for i 1 2 L testing a loop segment 30 f f 0 01 31 Re Re 0 01 32 end 33 34 Comment line here 35 36 L 5 Output FINAL VALUES OF VARIABLES 1 05000000000000 v 1 700000000000000E 003 1 5 00000000000000 mu 8 500000000000001E 005 re 52 5800000000000 f 1 29904761904762 i 9 00000000000000 rho 108 Sample Problem 4 This script is used to illustrate the use of functions and subroutines In addition expressions within subprograms are also illustrated Model Re RUL u f if Re lt 2500 Re f 16Re if Re 22500 if Re if Re 0 06Re if L 5 0 06Re if L 4 f 0 01 loop i 1to2L f f Re Re 0 01 Input File eguations6d isc program rho v L mu Re f i as real Initial Values rho 1 05 v 17 0e 4 L 5 mu 8 5E 5 Models Re Reynolds rho v L mu mu
20. code The first and last lines define a program that has the name main The second line declares a real variable x The third line evaluates the expression 2 3 and assigns the result to x An output statement is not required since iSCRIPT will automatically output the declared variable 4 2 2 Without a Program Structure Follow the same steps as in 4 2 1 However in the current case omit steps 3 and 6 The resulting program should be as shown in Figure 4 3 29 iSCRIPT Editor C iScript SampleScripts Math Example 4 1 math isc C File Edit Search View Document Project Tools Macros Window Help Deihew Be 648 e8 2 TAn TEA w yo amp math isc K as real a Ln1 Ch3 ASCII DOS READ REC COL OVR Figure 4 3 iSCRIPT code for the computation of 2 3 without a program structure Table 4 1 gives a few sample mathematical expressions and their corresponding iSCRIPT codes Math Expression iSCRIPT Code x as real x 2 5 1 v3 x 2 5 1 3 0 0 5 x 2 x y as real y sin xe In x 1 logo x x 2 y sin x exp x log xt 1 log10 x 12 0 3 107 25 A 2 2 B 2 2 C 2 2 as real A 1 2 0 3 0 7 2 5 gal ee B 5 0 7 7 12 5 2 3 12 5 2 3 C A B C AB Table 4 1 Sample mathematical expressions and their iSCRIPT equivalents 26 4 3 Another Simple Problem Aircraft Drag Calculation Let s consider another simple arithmetic problem involving ai
21. component previously declared with the CreateComponent command 57 Heat RejectionPemo isc program main l Declare the Tower Pump Condenser subsystems and their variables 2 Declare the HR sys system and its variables 3 Assign the input value of heat flux of the Condenser 4 Declare the objective variable HR sys C 5 Declare the optimization variables HR_sys w and Tower 6 Declare Cooling Tower component is a subcomponet of Heat Rejection System 7 Start to optimization end program subroutine HR sys 8 Execute the Tower component 9 Execute the Pump component 10 Execute the Condenser component 11 Calculate the total system cost Eq 5 1 end subroutine subroutine Tower 12 Get the heat flux of Tower by energy balacing Eqn 5 5 13 Calculate the temperature of water out of Tower Eq 5 6 14 Calculate the cost of the Tower Eq 5 2 end subroutine subroutine Pump 15 Calculate the cost of Pump Eq 5 3 end subroutine subroutine Condenser 16 The temperature of entering water of Condenser is the temperature of leaving water of Tower Eq 5 7 17 Calculate the cost of Condenser Eqn 5 4 end subroutine Figure 6 3 Program outline for Problem 6 2 The program outline is shown in Figure 6 3 above Compared with Problem 5 3 we only need to make the following changes in the main program Indicate the objective variables using AddObjective Indi
22. contain a main program which is described in 2 4 2 An example of a project file is shown below In the example below no path information is included as all the iSCRIPT files associated with the solution and listed in the project file are contained in the same folder project OLS_pumps isc OLS _lines isc OLS HX ise OLS exergy isc Ln7 Ch5 Figure 2 1 Sample project file 2 5 Running Sample Problems Sample problems are provided for you in the SampleScripts subdirectory of the folder in which you installed iSCRIPT They can be opened edited and run in iSCRIPT Editor environment In this manual several examples are provided and described in detail in Chapters 4 and 5 The procedure for launching the editing environment is described below 12 1 Open iSCRIPT Editor 2 IniSCRIPT Editor open the iSCRIPT project file HeatRejection ipr located in the SampleScripts HeatRejection subfolder of the isCRIPT installation folder iSCRIPT Editor Text1 Open Template Close Close All Reload Reload As Save Save As Save All Print Print Preview Print Setup FIP Recent Files Exit Open an existing document iSCRIPT Editor Text1 D Fie Edt Search peec cemas AIBA vi enh 2a Alt Shift 0 Alt Shift 5 x of 1 ASCI DOS READ REC COL OVA A Dens a S Text ASR SBS 02 DHHS SOBA WY eur 278 Look in HeatRejection gt e ef EA QExamples 3 Pum
23. execute PS exe if Girun 1 then call write imyoout Executable did not run or could not be located end if call write imyoout My current directory is currentdirectory changedirectory D alabi call write imyoout My new directory is currentdirectory changedirectory call write imyoout My final directory is currentdirectory call write imyoout call fprintfGimyoout 131 call write imyoout rho rho T T P P mu mu call rewind imyopen call read imyopen var2 call close imyopen call close imyoout Instructions 1 Run the above script 2 Examine the output files and compare with the input I O calls in the script as well as the output of Sample Problem 1 132 Sample Problem 19 Executing an iSCRIPT program in Parallel Any iSCRIPT program that can run on a single processor computer can execute in parallel in a multi processor environment However this example illustrates the use of iSCRIPT in optimizing a problem is parallel This sample problem file contains a main program and a subroutine which evaluates a model The model is the Rastrigin equation as shown below f x 20 x5 10 cos 27x cos 27 This function has several local minima making it difficult for a gradient based procedure to capture the actual minimum without the benefit of a good starting or guess value The actual minimum value is 0 and occurs at the values of x1 x2 0 0 G
24. in parallel in a multi processor environment However currently only the optimization portion of the program gains from the parallel environment This means that programs without any optimization calls will simply execute multiply on all the processors The system requirement for running iSCRIPT in parallel is described in the next section while procedures for running iSCRIPT programs in parallel is described in Section 6 2 7 1 System Requirements The iSCRIPT executable imparted with automatic parallel features and algorithm is named iscript_mp exe This program must be run on a computer or network of computers on which MPI has been set up The complete requirements are as follows iSCRIPT parallel executable iscript_mp exe which is available from an iSCRIPT installation Computer or network of computers on which MPI has been set up Ideally this should be a multi processor environment but MPI is also able to work on a single processor installation spawning virtual processes simulating a multi processor environment Each computer or the single computer simulating a parallel environment should be at least a Pentium PC running at a speed of at least 1GHz with at least 128MB ram and 200MB free disk space 7 2 Running an iSCRIPT Program in Parallel The syntax for running an iSCRIPT program in parallel is as follows 62 mpirun np iscript_mp exe or mpirun np iscript_mp exe program isc or mpirun n
25. mpich 2 3 Installing the Required Software 1 To install iSCRIPT e Double click the installation file T Message This program will install SCRIPT 2 0 Beta into your system Copyright c 2004 2007 TTC Technologies Inc info ttctech com 2010 20 Don t Install Help e Click Install Target Directory r Message eo Please enter the directory where the files of iSCRIPT will be copied to Cancel To accept the default press enter Help C MISCRIPT Browse e Select where to install iSCRIPT and click OK iSCRIPT 2 0 Beta Decompressing From C Documents and To C iSCRIPT iscripteditor exe Program Group Message OK s7 We are about to create a program group ar to hold the icons for iSCRIPT Please enter a name for it Cancel Help iSCRIPT 2 0 e Click OK Installation Completed Message ea Congratulations T iSCRIPT 2 0 Beta has been installed successfully e Click OK 2 Install MPICH Optional only needed for parallel calculation e Double click the installation file WinZip Self Extractor mpich nt 1 2 5 exe Unzip and start the setup utility to install MPICH on this machine You will be able to cancel later Si Cancel About e Click Setup Welcome to the MPICH Setup program This program will install MPICH on your computer It is strongly recommended tha
26. of the read keyword except when no argument is provided empty read arg1 arg2 argN are strings variables or arrays Strings must be enclosed in single quotes A is a variable or array 89 e When specified isize refers to the total number of elements that should be read When isize exceeds the size of the array the size of the array is used e Ifan end of file occurs during a read the command returns and the program resumes An internal flag is set which may be queried using the eof command see Section A 8 6 A 8 4 Writing to a File or the Screen The syntax to write to an open file or the standard output usually the screen is shown below call write unit format arg1 arg2 argN call write unit format arg1 arg2 argN or isize write unit format arg1 arg2 argN isize fprintf unit format arg1 arg2 argN The above rules govern the process of calling the write command e unit is an integer between 10 and 100 representing the handle for the open file No unit specified units 1 2 5 or 6 refers to the keyboard e format is a string representing the read format The format string may be omitted simply provide an empty or an used instead The format string when read is currently ignored but is accepted for compatibility with future versions of iSCRIPT The MATLAB format specifiers are accepted e fprin
27. optimization variables and they belong to different components Therefore we need to indicate the relationship between the Heat Rejection System and the Cooling Tower component This is done by indicating which components belong to each subsystem In a multi subsystem environment several subsystems may be present in the entire system and the subsystem composition must be indicated Otherwise it would be impossible to determine which component belongs to which subsystem In the current example there is only one subsystem thus this is also the system The subsystem composition is indicated using the AddSubComponent command The component subsystem system relationship in isCRIPT is summarized below Every iSCRIPT project is assumed to contain one system a system is automatically created per project All subsystems in a project automatically belong to the system If there is only one subsystem then this subsystem comprises the system The component subsystem relationship or hierarchy is formalized using the AddSubComponent command The format for the AddSubComponent command is shown below AddSubsystem component subcomponent component The component consisting of other components or subsystem a string limited to 24 characters This component must be a component previously declared with the CreateComponent command subcomponent A component to be identified as a subcomponent of a component This component must be a
28. performed the traditional way without using the ILGO procedure This will usually result in nested optimization loop In this case subsystems do not need to be indicated and inter component coupling are simply contained within the models Optimization is called for each component as well as for the sub system which contains components for which optimizations are called Optimization is also called for an overall system component which is created and contains every other component For the system in Figure 6 the calls will be as shown below Componenti model Component2 model Componentn m model equations equations equations Component1 Optimize Component1 Optimize Componentn m Optimize Componentx model Componenty model Componentz model Subsystem 1 Subsystem 2 Subsystem L equations equations equations Componenti Execute Componentn 1 Execute Componentn m k 1 Execute Component2 Execute Componentn m Execute Componentn m k 2 Execute Componentn Execute Componenty Optimize Componentz Optimize Componentx Optimize Componenta model equations Componentx Execute Componenty Execute 102 Componentz Execute Componenta Optimize Note that for any fairly detailed system the above nested optimization arrangement will be very time consuming even if the first level is eliminated and the subsystems are optimized integrally using the free variables from their components B 2 10 Optimization Genetic Algorithm The geneti
29. the iSCRIPT working directory at the time of the call see Section 7 3 The filename must be enclosed in single quotes e commandline_argument is an optional string containing arguments that should be passed to the executable at runtime e The execute command may be used to run any executable at runtime including Microsoft Excel other CFD programs and MATLAB using MATLAB s component compile MCC tool 66 e Combining the ability to run an external executable with the input output procedures in Section 4 8 allows iSCRIPT to write input files for an external program run the program and read the output This allows engineering models implemented in other environments to easily be integrated with an isSCRIPT solution See sample problem 17 for an example solution integrating an aircraft engine model program written in MATLAB 8 3 Setting Changing the Working Directory iSCRIPT provides commands for changing the execution directory in case a third party executable does not reside in the same folder as an iSCRIPT program that wishes to run the executable The syntax to accomplish this is call changedirectory path or ivar changedirectory path The above rules govern the process of calling the changedirectory command e path is a string representing the directory to change to When not provided the command reverts to the executing iSCRIPT directory See sample problem 18 Path must be enclosed in single quotes
30. them in an expression in lieu of a variable The statement var functionname arg1 arg2 argN where arg1 arg2 argN are the actual arguments to the function assigns the value of the function to the variable var To solve this problem it is convenient to create a function get_area to calculate the surface area of the cylinder and a subroutine get_volume to calculate the volume of 33 cylinder The complete code is shown in Figure 4 8 The calls to the function get_area and the subroutine get_volume are in lines 9 and 11 of the main program Lines 15 to 24 define the get_area function Note that the type of the function is declared on line 17 get_area as real This is required when writing a function in iSCRIPT Lines 26 to 35 define the get_volume subroutine Note that the return value of volume must be an argument of the subroutine This value will be computed when the program is executed and the computed value will be returned to the main program Note For the subroutine function the names of the dummy arguments and the actual arguments do not have to be same but they are required to have the same type IniSCRIPT at runtime the values of the actual arguments if modified within the subroutine are returned to the calling program or subprogram After control is returned to the calling routine the local variables in the subroutine or function are automatically freed CircularCylinde
31. to be passed into subprograms if their values are intended to change in any subprogram A 6 Other Program Flow Structure For compatibility with other engineering programming tools iSCRIPT supports additional syntax including the break and continue keywords as well as labels and go to statements A 6 1 Break This keyword is used only for scripts indicated as MATLAB source The break statement terminates the execution of a loop segment In nested loops the break statement only exits the loop within which it occurs Examples for ii expression1 expression2 break end for for ii expression1 expression2 if expression3 then break endif 85 end for Both examples above cause the premature termination of the for loop on encountering the break statement In the second case the termination only occurs on the condition of expressions A 6 2 Continue This keyword is used only as a place holder For instance the continue keyword may be used to establish a label The presence of the continue keyword has no effect whatsoever in a scripting segment Examples are included in A 6 3 A 6 3 Go to and labels The goto statement is used to influence program flow This statement is used in conjunction with a label statement The syntax is shown below goto label label The rules governing goto and label statements are as follows e label is an alphanumeric word defined according to the rules for naming variables as descr
32. will narrow this space to from L M to MAX L x A MIN M x A where x is the default value of x a computed initial value for x or the value of x after a prior optimization step This parameter may be used to further reduce the search space after a prior optimization step or after an initial computation designed to obtain a good initial estimate narrowing the range of the optimum value of x to speed up the optimization process 52 Note The objective and optimization variable are previously declared component variables There can be more than one optimization variable Note that the fewer optimization variables there are the faster the optimization will complete Also the smaller the search space of the optimization variable the faster the optimization will complete After both the objective variable and the optimization variables are declared by AddObjective and AddVarObjective statements the optimization calculation is executed by using the statement Component Optimize 6 2 Optimization of a System with a Single Component Let us continue to use the Circular Cylinder problem for the example The model used for the components of the system are very rudimentary but serve to illustrate how iSCRIPT can be used to model engineering systems The problem is revised as follows Problem 6 1 Minimize the Surface Area of a Circular Cylinder with a Fixed Value of Volume of 800 m
33. 00000000E 002 nsampsize 70 Evaluations and range 256 Population limit initial trials 140 TRYING variables O 0 1000E 01 0 1000E 04 OPTIMIZING INITIAL CONTINUING 2000 OPTIMIZING CONTINUING 5 Number of individuals after optimization 140 Individual 1 Objectives 0 4880652E 05 0 4880652E 05 Individual 1 Variables 0 2042933E 03 0 1566025E 00 Individual 2 Objectives 0 5050623E 05 0 5050623E 05 Individual 2 Variables 0 2304584E 03 0 1949618E 00 Individual 3 Objectives 0 5443953E 05 0 5443953E 05 Individual 3 Variables 0 3151042E 03 0 2168838E 00 Individual 4 Objectives 0 4413595E 05 0 4413595E 05 Individual 4 Variables 0 1926272E 03 0 1135623E 01 Individual 5 Objectives 0 4964144E 05 0 4964144E 05 Individual 5 Variables 0 2596472E 03 0 1984522E 00 Individual 6 Objectives 0 5943523E 05 0 5943523E 05 Individual 6 Variables 0 3578200E 03 0 1939299E 00 Individual 7 Objectives 0 4135938E 05 0 4135938E 05 Individual 7 Variables 0 2319260E 03 0 4035045E 01 Individual 8 Objectives 0 5544111E 05 0 5544111E 05 Individual 8 Variables 0 1653570E 03 0 1249507E 00 Individual 9 Objectives 0 5995108E 05 0 5995108E 05 Individual 9 Variables 0 3634436E 03 0 1547490E 00 Individual 10 Objectives 0 4292897E 05 0 4292897E 05 Individual 10 Variables 0 2460449E 03 0 8659816E 01 Individual 11 Objectives 0 4585632E 05 0 4585632E 05 Individual 11 Variables 0 2664919E 03 0 1042430E 00 Figure 6 3 Sample optimize txt output file 65 8 Interface with O
34. 04 134 1 Introduction 1 1 Introduction to Industrial Design Optimization iSCRIPT was developed primarily for performance calculations and design optimization of engineering systems with a focus on large systems As a result iSCRIPT was designed to include tools for modeling and optimizing large scale systems It has provisions for modeling a system in a decomposed fashion and automatically executes in parallel taking advantage of multi processor computational resources iSCRIPT also contains novel optimization procedures which allow for the optimization of decomposed systems in an efficient manner To facilitate detailed modeling of engineering components and systems iSCRIPT has the full features of a programming language comparable to those available in traditional programming environments such as FORTRAN A common technique for carrying out a multi level design optimization can be divided into three tasks as shown in Figure 1 1 below Figure 1 1 Tasks in a multi level design optimization analysis 1 1 1 Decomposition In physical decomposition the system is divided up into physically interacting subsystems each possessing a certain degree of autonomy but depending on other subsystems via a number of coupling variables Disciplinary decomposition divides the system along the lines of different disciplines such as thermodynamics economics aerodynamics etc Conceptual decomposition breaks down the system according to the
35. 24 characters as defined in A 10 1 B 2 5 Procedures for Performing an ILGO Optimization In addition to the information required for optimizing components sub systems require the following information An indication of the subsystems in the project or system each project is automatically assumed to represent one system The list of components that make up the sub system as indicated in Section 11 4 An indication of the coupling variables between sub systems The procedure for providing the above information is described in subsequent sections B 2 6 Indicating a Subsystem A subsystem may be indicated using the following command AddSubsystem component component The component to flag as a subsystem a string limited to 24 characters This component must be the component to be optimized or a sub component of it This component must be a component previously declared with the CreateComponent command B 2 7 Indicating Inter Component Subsystem Coupling A coupling between two components subs systems as follows AddCoupling component1 variable1 component2 variable2 componenti The component to which the variable belongs a string limited to 24 characters This component must be the component to be 100 optimized or a sub component of it This component must be a component previously declared with the CreateComponent command variable Name of the variable a string limited to at most 24 c
36. 5 Save the file Step 6 Select Tools gt Run Current Script Project File from the menu to run the iSCRIPT code Step 7 Find the solution on the screen or in the file outputscript txt The first part of the main program should consist of the component declaration of the Cylinder component using a CreateComponent statement and the declaration of its four component variables using a CreateVariable statement Then the cylinder radius and height should be assigned The Cylinder component may be executed to calculate the surface area and volume by using Cylinder Execute Beside the main program an executable subroutine which has the same name Cylinder should be created for the Cylinder component to compute the surface area and volume Figure 5 1 shows the program outline for this problem You may replace the comment statements of steps 1 through 5 with your own iSCRIPT code segment to complete the program A sample completed program is shown in Figure 5 2 CircularCylinderDemo isc Program main 1 Create the Cylinder component 2 Create the four component varialbes of the Cylinder component 3 Assign the input value for the radius and height 4 Execute the Cylinder component end program subroutine Cylinder 5 Calculate the surface area and volume end subroutine Figure 5 1 Program outline for Problem 5 1 We will now compare the script in Figure 5 2 with the program outline in Figure 5 1
37. 5E 4 alt 1 7 mu 1 61988E 7 T T T 4 66143E 4 T T 7 25242EF 1 T 4 20454 1e 7 8 rho P 287 T Output The output is presented below Only the print out of the final value of all variables is presented A hand calculation may be used to confirm the accuracy of the parsed results FINAL VALUES OF VARIABLES alt 914 4000 t 311 9852 p 90830 66 mu 1 9001649E 05 rho 1 014417 alt_1 3000 000 106 Sample Problem 2 This script is used to illustrate the use of if statements nested if statements and if statements with multiple brackets and single variable expression as the condition Model _ PUL u Ja if Re lt 2500 Re f 16Re if Re 2500 f 16Re if Re ANDL 5 Re Input File eguations4 isc 1 rho v L mu Re f 2 rho 1 05 3 v 17 0e 4 4 Le 5 5 mu 8 5E 5 6 Rez rho v L mu 7 if Re lt 2500 then 8 f 64 0 Re 9 end 10 if Re gt 2500 then 11 f 16 0 Re 0 4 12 end 13 if Re then 14 if L 5 then 15 f 0 06 Re 0 4 16 end if 17 end 18 19 L 4 Output FINAL VALUES OF VARIABLES tho 1 050000 v 1 7000000E 03 l 4 000000 mu 8 5000000E 05 re 105 0000 f 9 3255732E 03 107 Sample Problem 3 This script is used to illustrate the use of do and for loops For loops are treated equivalently as do loops for compatibility with MATLAB syntax Comments and in script documentation are also illustrated Model Input File eqguations6 isc
38. 81859 Double Variables 0 125 fe Micr n Debug File Edit File Edit View Window Help De X Bo 2 curent directory 4 2 3 4 6 6 7 8 9 2 1 1 1 2 13 1 1 2 4 1 0 2 1 2 5 1 1 y x b x 2 y x x 2 sin y Ln 4 2 1 l 1 2 1 410 212 51 41 6667 35 0000 36 3333 47 6667 54 0000 49 3333 haba ane Die 50 3333 43 0000 61 6667 2 sin y 4 2 3 4 6 6 f i i ip 2 i 410 212 551 ytx btx 2ty x 5 8186 3 6829 4 6829 5 6829 7 8186 7 6829 8 6829 9 6829 10 8186 Nn Oo Ww xX N DGM 2 sin y 8 22 06 7 51 AM 126 Sample Problem 15 MATLAB Compatibility This script was used to illustrate compatibility with MATLAB function features including the management of function arguments and the use of multiple output arguments The model is the similar to that in Sample Problem 6 Input File matlabal4 isc program rho v L mi Re f i rtestl rtest2 as real initest3 as integer Initial values rho 1 05 v 17 08 4 L 5 SE initest3 2 Models Re Reynolds rho v 4 0 mu m 8 56 5 rtestl rtest2 initest3 Computef Re f Comment Tine here Cs end program function results Reynolds rho u L mu rho U L my results i as real iComputef as integer results rho u L m icomputef 3 4 u icomputef 12 2 end function function outputl output2
39. CONTINUING OPTIMIZING CONTINUING 560 Number of individuals after optimization 140 Individual 1 Objectives 0 4880652E 05 0 4880652E 05 Individual 1 Variables 0 2042933E 03 0 1566025E 00 Individual 2 Objectives 0 5050623E 05 0 5050623E 05 Individual 2 Variables 0 2304584E 03 0 1949618E 00 Individual 3 Objectives 0 5443953E 05 Individual 3 Variables 0 3151042E 03 Individual 4 Objectives 0 4413595E 05 Individual 4 Variables 0 1926272E 03 Individual 5 Objectives 0 4964144E 05 Individual 5 Variables 0 2596472E 03 0 1984522E 00 Individual 6 Objectives 0 5943523E 05 0 5943523E 05 o o 0 5443953E 05 o o o o o o Individual 6 Variables 0 3578200E 03 0 o o o o o o o o o o 2168838E 00 4413595E 05 1135623E 01 4964144E 05 Individual 7 Objectives 0 4135938E 05 Individual 7 Variables 0 2319260E 03 Individual 8 Objectives 0 5544111E 05 Individual 8 Variables 0 1653570E 03 Individual 9 Objectives 0 5995108E 05 Individual 9 Variables 0 3634436E 03 Individual 10 Objectives 0 4292897E 05 Individual 10 Variables 0 2460449 03 Individual 11 Objectives 0 4585632E 05 Individual 11 Variables 0 2664919E 03 0 1042430E 00 Figure 6 5 Sample optimize txt output file The sample problem described in this section is a simple one In this problem the overall objective function can be expressed as a single equation 14x 10 C f w A 800A 0 005w 270 6 3 os f w A aT 6 3 Therefore the advantages of dec
40. INAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 0 Real Variables 7 rho 1 050000 v 12 20000 1 5 000000 mu 8 5000000E 05 re 105 0700 f 7 9325572E 02 i 8 0000000E 00 114 Sample Problem 7 This script is used to illustrate the use of array declaration and the use of arrays within expressions This program is part of a model which computes the aerodynamic characteristics of an aircraft fuselage This program is also useful for demonstrating the use of arrays in an iSCRIPT code Model j 0 Sy 335 V 856 blogic 2 2 2jj 1 ik blog ic 2 2 2 j _ PUL fuselage Re I 1 053 RE wutoff_ sub asal 1 053 Re aaea M S C 0 454 fuselage flog Re sheer h 0 144M Re cutoff _sup Amax wee 3 83 294 54 2 55 2941S l Pcie T 0 3048d ign FF p selage 1 aa f on 400 S wer shoe uselage FF fuselage Cy C dO fusetage jiseage S ref 115 Input File eguations9 isc 1 program main 2 3 icount ik jj kk as integer 4 5 Sref Wo k Swet_fuselage alt mu rho R_fuselage R_cutoff_sub Cf_fuselage Amax d f_fuselage R_cutoff_sup as real 6 FF_fuselage Cdo_fuselage V M as real 7 8 iunit ifile i j as integer 9 10 blogic 3 2 brun as logical 11 12 lexist lrun as logical 13 14 il as integer 2 15 j3 j4 as integer 2 16 17 rlong rlonger as double 18 rnumber rnumb rexp as double 21 ModelEquations 22 function Cdo_sup
41. RIPT system project Consequently iSCRIPT can accommodate virtually any decomposition procedure as the components may be models of physical components conceptual segments of a system or a disciplinary subdivision of a system iSCRIPT also contains built in procedures for optimizing a system made up of decomposed subsystems using the iterative local global optimization ILGO procedure As previously mentioned iSCRIPT was designed to automatically execute in a parallel environment taking advantage of multi processor facilities for faster turn around on simulation tasks To provide system of systems features iSCRIPT can execute third party software potentially providing access to models available on other software products iSCRIPT is also compatible with some engineering modeling environments such as FORTRAN and MATLAB Programs developed in these environments can with minimum modification be run in iSCRIPT or included as part of an iSCRIPT system model 1 4 Conventions Used in this Manual The following conventions have been adopted in this manual e Arial font will be used for script segments iSCRIPT keywords and commands that should be typed on the keyboard as input For example if Burner T gt 2700 then CreateComponent name description Segments of a syntax enclosed within square brackets are optional For instance given the above syntax template for the CreateComponent command the following two instances of
42. al procedures in Section 3 2 The outline of the script program is shown in Figure 5 10 The problem has one system HR_sys and three components Tower Pump and Condenser Both the system and the components and their variables need to be declared Each component must be simulated in its own component subroutine The component and component variable declaration are done in the main program 46 Heat RejectionDemoZ isc program main 1 Declare the Tower Pump Condenser components and their variables 2 Declare the HR sys system and its variables 3 Declare the objective variable HR sys C 4 Assign the input value of Tower A and HR_sys w 5 Execute HR sys system end program subroutine HR_sys 6 Execute the Tower component 7 Execute the Pump component 8 Execute the Condenser component 9 Calculate the total system cost Eqn 5 1 end subroutine subroutine Tower 10 Get the heat flux of Tower by energy balance Eqn 5 5 11l Calculate the temperature of water out of Tower Eq 5 6 12 Calculate the cost of the Tower Eq 5 2 end subroutine subroutine Pump 13 Calculate the cost of Pump Eq 5 3 end subroutine subroutine Condenser 14 The temperature of entering water of Condenser is the temperature of leaving water of Tower Eq 5 7 15 Calculate the cost of Condenser Eq 5 4 end subroutine Figure 5 10 Program outline for Problem 5 5 The complete code is written in six f
43. and expert systems are more computationally intensive but they are not as prone to local optima are more tolerant to arbitrary initial guess and can be used for mixed integer problems In aerospace engineering the variables include integers Boolean and continuous variables GAs are utilized to isolate the optima while the gradient based method can be combined with the GA to speed up the climb to the peak once all integer variables are set In other words a combination of several optimization procedures is typically used for a complex problem 1 2 Why We Developed iSCRIPT iSCRIPT was developed mainly because existing tools for engineering design were not explicitly developed for that purpose and were therefore unable to take advantage of the inherent structures that are present in engineering problems Engineering context includes engineering units material or physical limits constraints or the natural association of variables with components or system and the association of components with systems or subsystems Engineers must fill this gap by writing codes to convert variables between different engineering units and procedures and to coordinate the association of components within a system This makes the development of detailed large systems very unwieldy and error prone With improved computational power and little additional work engineers can now afford to carry out performance analysis and design optimization to increase
44. as real i as integer Lines 3 through 6 declare the variables that are used in the program Accepted variable types include logical short integer real and double An exhaustive list can be found in the iSCRIPT Language Reference in Appendix A Also notice that variables V_res and drag_res are declared as real arrays intended to accommodate 11 real values 8 drag 20000 9 rho 1 0e 6 10 100 0 4470 11 A 1 0 Lines 8 through 11 are called assignment statements These assign specific values to the variables 13 Cd 2 drag rho V 2 Line 13 represents the calculation of the drag coefficient 15 do i 1 11 16 V_res i 200 0 4470 10 i 1 I7 drag_res i Cd rho V res ij 2 a 2 18 end do Lines 15 through 18 represent the calculation of the drag for 11 speeds ranging from 0 to 200 mph Notice the conversion of the units of speed to SI units m s by multiplying by the factor 0 4470 31 4 4 Another Arithmetic Problem Circular Cylinder Geometry We choose the circular cylinder surface area and volume calculation problem as another example to illustrate the use of isCRIPT for a simple arithmetic problem In this problem we will describe how to use the subroutine and function structures in iSCRIPT The problem is described as follows Problem 4 3 Calculate the Surface Area and Volume of a Circular Cylinder Figure 4 7 Physical problem and its primary variables radius rand height
45. as the project file Figure 5 3 shows an example of a project file Note that the subfolders PS OLS CHS ECS VCPAOS FLS and AFS are subfolders of the project file ata ipr ata ipr project main isc PS PS ise OLS OLS isc CHS CHS isc ECS ECS isc VCPAOS VCPAOS isc FLS FLS isc AFS AFS isc Figure 5 3 An example of a project file To run a problem with a project file simply open the project file and select Tool gt ISCRIPT from the menu to run the whole project Note Ina project there must be a program in a script file which will be the starting point of the program execution Jfa script file is in the same folder as the project file the directory or path information may be neglected e g main isc in Figure 5 3 The order of the script files is not important 5 5 An Example of a Program Developed in Several Script Files Here again we use the Circular Cylinder problem as an example Problem 5 2 Solve Problem 4 3 Using Several iSCRIPT Files The code developed in Problem 4 3 has one main program one function and one subroutine These three program structures can each be written in a separate file Thus together with a project file this project will now be written in 4 files These file are listed in Figures 5 4 through 5 7 42 Cyliner ipr project Cylinder isc rea isc Volume isc Figure 5 4 Project file Cylinder i
46. ays named according to conventions described in Section A 1 and are called the actual arguments to the subprogram The names of the actual arguments may be different from those of the dummy arguments in Section A 5 2 However the type and array sizes must match if the actual and dummy arguments are arrays The program or subprogram within which the above statement is placed is referred to as the calling program or subprogram The actual arguments to the subprogram must be declared in addition to any other variable declared within the calling program or subprogram At runtime the values of the actual arguments if modified within the subroutine is returned to the calling program or subprogram Function Structure function functionname arg1 arg2 argN variable declaration statements end function The following rules govern the use of the function structure statement A 5 5 functionname is the name of the function and must be named according to variable naming conventions described in Section A 1 variable declaration statements represents several lines of variable declaration statements as described in Section A 1 arg1 arg2 argN are variables or arrays named according to conventions described in Section 4 1 and are called the dummy arguments to the subprogram The dummy arguments to the subprogram must be declared in addition to any other variable declared within variable declaration statements functionname must
47. be declared in addition to any other variable declared within variable declaration statements functionname may be declared as an array represents one or more lines of scripting language segments and is referred to as the body of the subprogram endfunction may also be used instead of end function Calling a Function A function may be called within a program other subroutine or function A function may also be called recursively Functions may be called simply using them in an expression in lieu of a variable or array var functionname arg1 arg2 argN 83 Examples Function Calls Actual Function Si Function Reynolds_No r V L Re Reynolds_No rho U L mu mu A r V L mu as real Re_is Reynolds_No rho U L Reynolds No r V L mu mu gm End Function The following rules govern the process of calling a function e varis the name of a variable or array named according to conventions described in Section A 1 e functionname is the name of the subroutine and must be the same as that used in the function keyword in Section A 5 2 e argi arg2 argN are variables or arrays named according to conventions described in section A 1 and are called the actual arguments to the function e The names of the actual arguments may be different from those of the dummy arguments in Section A 5 2 However the type and array sizes must match if the actual and dummy arguments are arrays e The program or subprogram wi
48. be only one program in a model of a component Subroutine Structure subroutine subroutinename arg1 arg2 argN variable declaration statements end subroutine The following rules govern the use of the subroutine structure statements A 5 3 subroutinename is the name of the subroutine and must be named according to variable naming conventions described in Section A 1 2 variable declaration statements represents several lines of variable declaration statements as described in Section A 1 1 arg1 arg2 argN are variables or arrays named according to conventions described in Section A 1 2 and are called the dummy arguments to the subprogram The dummy arguments to the subprogram must be declared in addition to any other variable declared within variable declaration statements represents one or more lines of scripting language segments and is referred to as the body of the subprogram endsubroutine may also be used instead of end subroutine Calling a Subroutine A subroutine may be called within a program other subroutine or function A subroutine may also be called recursively Subroutines may be called using the call keyword call subroutinename arg1 arg2 argN The following rules govern the process of calling a subroutine subroutinename is the name of the subroutine and must be the same as that used in the subroutine keyword in A 5 2 82 A 5 4 arg1 arg2 argN are variables or arr
49. c algorithm used in iSCRIPT is based on a modification of the method of Geoff Leyland The principles used are as follows Generate an initial population by sampling sparsely over the combinatorial search space of all variables combined Improve the population over a number of generations by combining individuals within the population Combination is created using the following operators o Selection of combining or mating individuals based on a random selection process weighted to more likely select individuals at the top o Combining the individuals using a blended function of the free variables o Interrupt the process at a low frequency using a mutation operator to ensure that the algorithm does not settle into a non optimal subspace o Replace only the bottom half of every generation after every combination cycle Inherent in the above procedure is a thinning strategy that limits the population size to a specific value for practical purposes During both the initial population and improvement phases new individuals are inserted into the sorted population such that worse individuals drop off once the population size is at a limit The values of the population limit initial sampling size number of generations and the mutation frequency are variables that affect the genetic algorithm Default values have been set for these parameters in iSCRIPT but can be modified as described in the next section B 2 11 Optimization Paramete
50. cale factor Some examples of legal numbers are 7 82 0 00007 e 7 4382259 1 60210e 20 9 1345 e14 All numbers are stored internally using the double type described in Section 2 1 and as specified by the IEEE floating point standard A 1 4 Assignment Operator The symbol is used as the assignment operator var1 1 002 A 1 5 Arithmetic Operators Arithmetic operators include Operator Function 5 Subtraction Subtracts a variable number or expression on the right from a variable number or expression on the left of the operator Addition Adds a variable number or expression on the right to a variable number or expression on the left of the operator jj Multiplication Multiplies a variable number or expression on the right to a variable number or expression on the left of the operator Division Divides a variable number or expression on the left by a variable number or expression on the left of the operator 75 Power Raises a variable number or expression to the left to a power defined by a variable number or expression on the right Brackets Independently evaluate the value contained within the brackets The precedence of the operators is as shown from top to bottom i e are computed before when they are in the same expression A 1 6 Relational Operators Relational operators include Operator lt Function Less than Compares a variabl
51. calling the open command e unit is an integer between 10 and 100 provided as a handle for opening the file This handle should be used when reading from or writing to the file When the unit is an output iSCRIPT opens the file on an available unit and supplies the unit as the function output e filename is a string representing the name of the file to be opened and must be in accordance with the file naming rules on the operating system Filename must be enclosed in single quotes e fopen may be used instead of the open keyword e Permission_mode is one of the options specified in the table below Permission Mode Specifiers Permission_mode Description Text Mode Open file for reading default Open file or create new file for writing discard existing contents if any Open file or create new file for writing append data to the end of the file Open file for reading and writing Open file or create new file for reading and writing discard existing contents if any Open file or create new file for reading and writing append data to the end of the file Binary Mode Open file for reading default Open file or create new file for writing discard existing contents if any Open file or create new file for writing append data to the end of the file Open file for reading and writing Open file or create new file for reading and writing discard existing contents if any Open file or create new file for reading and
52. cate the optimization variables using AddVarObjective Start the optimization calculation using HR_sys Optimize 58 main isc This is the main program for the heat rejection system optimization program main 1 Create the subsystem Create the Tower subsystem and its variables CreateComponent Tower the cooling tower CreateVariable Tower C double 0 0 1 0E9 0 0 0 0 CreateVariable Tower A double 0 0 500 0 10 0 105 0 m 2Z CreateVariable Tower Q double 0 0 1 0E9 0 0 0 0 W CreateVariable Tower Tout double 0 0 500 0 0 0 0 C Create the Pump subsystem and its variables CreateComponent Pump the pump CreateVariable Pmp C double 0 0 1 0E9 0 0 0 0 Create the Conmderser subsystem and its variables CreateComponent Condenser the condenser CreateVariable Comdenser C double 0 0 1 0E9 0 0 0 0 CreateVariable Condenser Q double 0 0 1 0E9 0 0 0 0 W CreateVariable Condenser T_in double 0 0 500 0 0 0 0 C 2 Create the systems Create the heat rejection system and its variables CreateComponent HR_sys Heat Rejection System CreateVariable HR sys C double 0 0 1 0E9 0 0 0 0 CreateVariable HR sys w double 0 0 500 0 10 0 220 0 kg s 3 optimization the system AddObjective HR_Sys C 0 AddVar Objective HR_ Sys w AddVar Objective Tower A AddSub Component HR_Sys Tower Optimize the System Global maxPopulation 2000 Global maxIn
53. d levels of fidelity iSCRIPT takes advantage of modern computer architecture by implementing procedures to automatically operate in a parallel environment without requiring the designer to explicitly parallelize models as would be necessary in a traditional programming environment One problem facing designers of large scale engineering systems is that the models for the design are not always available a single software or programming platform For instance engineers designing aircraft may sometimes call on the code WATE for the modeling of an engine separate from other models which may be implemented in other software products System modeling and design tools must allow systems to be built based on components some of which are modeled in a different software or programming platform iSCRIPT is being developed with this cross platform operability in mind 1 3 How We Want You to See ISCRIPT Although iSCRIPT can be used in the same manner as a traditional programming environment with all the features of a programming language including decision structures loop elements array variables and subprogram units it was designed primarily for performance analysis and design optimization of engineering systems Systems can be described in a building block manner as being composed of subsystems which in turn are composed of components Figure 1 2 Component Component Variables Figure 1 2 Elements of an iSC
54. different folder click Browse and select another folder You can choose not to install MPICH by clicking Cancel to exit Setup Destination Folder C Program Files MPICH Browse lt Back Cancel e Click Next Select Components Select the components you want to install clear the components you do not want to install Components runtime dlls Description This component contains the dlls necessary to run mpich applications It must be installed on each machine Space Required 11656 K Space Available 3198488 K lt Back Next gt Cancel e Click Next Start Copying Files Setup has enough information to start copying the program files If you want to review or change any settings click Back If you are satisfied with the settings click Next to begin copying files Current Settings Destination C Program Files MPICH lt Back Cancel e Click Next c program files mpich mpd bin mpirun exe 20 More information about the installation of MPICH can be found on the MPICH website 3 Installed Files After iSCRIPT is installed several folders and executable files are installed in the installation directory default is C SCRIPT These main files or folders are iSCRIPT Folder iscripteditor exe sysdes exe parser exe iscript_mp exe SampleScripts ata doc uninst folder folder folder folder 10 i
55. e number or expression on the right to an variable number or expression on the left of the operator Returns a value of 1 or true if the value on the left is less than that on the right Greater than Compares a variable number or expression on the right to an variable number or expression on the left of the operator Returns a value of 1 or true if the value on the left is greater than that on the right Less than or equal Compares a variable number or expression on the right to an variable number or expression on the left of the operator Returns a value of 1 or true if the value on the left is less than or equal that on the right Greater than or equal Compares a variable number or expression on the right to an variable number or expression on the left of the operator Returns a value of 1 or true if the value on the left is greater than or equal that on the right Equal Compares a variable number or expression on the right to an variable number or expression on the left of the operator Returns a value of 1 or true if the value on the left is equal that on the right Not equal Compares a variable number or expression on the right to an variable number or expression on the left of the operator Returns a value of 1 or true if the value on the left is not equal that on the right 76 A 1 7 Logical Operators Logical operators include Operator Function amp LOGICAL AND Returns a value of 1 or true if
56. e Type matlaba70 in the MATLAB command window e The output values of the matrix Zand are printed out 68 File Edit View Web Window Help D eS e d th Current Directory D alabiResearchifunction _parseriin_parseriSampleScripts f Workspace ax Command History gu ei Stade Bas x 4 2 3 4 6 6 7 8 9 1 0 0 Name Size By z 1 0 0 a 2 1 3 Ha S b y x ta Hr 3x3 x 4 2 37 4 6 6 7 8 9 Hx 3x3 Y l 0 0 Hay 3x3 a 2 1 3 Hz 3x3 be y x ta xX x 2 x 4 2 3 4 6 6 7 8 9 y 1 0 0 a 2 1 3 b y xj a x x42 6 29 07 8 12 AM matlaba70 Com Using Toolbox Path Cache Type help toolbox_path_cache for more info To get started select MATLAB Help from the Help menu gt gt matlaba70 41 6667 35 0000 36 3333 47 6667 54 0000 49 3333 50 3333 43 0000 61 6667 5 8186 4 6829 5 6829 7 6829 8 6829 10 8186 lt gt gt Workspace Current Directory Figure 7 1 MATLAB program showing results of running matlaba70 m script file 8 5 2 Running the program in iSCRIPT The iSCRIPT version of the same program is presented below Notice that the script is exactly the same except that the variables used are declared at the start of the program This is the most prominent modification that has to be made to MATLAB scripts to run them in iSCRIPT i x 3 3 y 3 3 b 3 3 as integer 2 3 3 a 3 3 as real x
57. e following rules govern the use of the if statement e expression1 expression2 and expression are valid expressions constructed as in Section 4 1 e represents one or more lines of scripting language segments which may include other if statements e else if may also be used instead of elseif e endif or end may also be used instead of end if e Itis not mandatory to have the elseif or else portions of the if statement e There is no limit to the number of elseif segments that may be included in an if structure e There can be only one else statement in an if structure e if statements may be nested as desired A 4 Loop Structure iSCRIPT uses the do for and while statements to implement the conditional execution of segments of a program The syntax for each type is described below 79 A 4 1 Do Loops do ii expression expression2 end do The following rules govern the use of the do statement A 4 2 ii is a variable declared as in Section 4 1 ii may be a short long real or double variable expression1 and expression2 are valid expressions constructed as in Section 4 1 represents one or more lines of scripting language segments which may include other do statements and is referred to as the body of the loop ii is incremented by 1 and the body of the loop executed until expression1 is greater than expression2 The body is not executed at all if expression is greater than expression2 at t
58. e the text form of the equations Example code X y as real x 2 y log x sin x exp x Save the file and run the file by select menu Tool gt Run Current Script Project File Find result on screen or in the file outputscript txt Figure 3 1 The basic procedure for solving an arithmetic and algebraic problem in iSCRIPT Figure 3 1 summarizes the basic procedure for solving an arithmetic and algebraic problem in iSCRIPT The steps are also listed below 16 Step 1 Open iSCRIPT Editor by selecting iSCRIPT Editor from the Windows Start menu Step 2 Create a new file in iSCRIPT Editor and save it as an isc file Step 3 Type the word program main in the first line of the file Step 4 Declare all the variables in the equations by typing the expressions like x as real x y as integer etc Variable types and their declaration in iSCRIPT can be found in Appendix A 1 1 Step 5 Type the equations in text form such as x 2 3 y log x sin x exp x2 etc The mathematical operations supported in iSCRIPT can be found in Appendix A 1 Step 6 Type the word end program in the last line of the file Step 7 Save the file Step 8 Select Tools gt Run Current Script Project File from the menu to run the iSCRIPT file Step 9 View the output on screen or in the file outputscript txt Note that Steps 3 and 6 may be omit
59. ective function and x7 x2 as variables to be optimized in achieving the minimum value of the objective function Lines 14 through 17 The second part of the program sets the optimization parameters Lines 22 27 and optimizes the component Line 29 Component Model The function or component is modeled in subroutine Rastrigin Lines 34 through 41 Variables are declared and initialized in Lines 33 through 39 of the program Equation 6 1 is implemented in Line 41 of the program Output Again the output file is outputscript txt The correct results were obtained in about 2 9 seconds on a Pentium workstation using only a population of 70 realizations Details of the optimization process are recorded in the file optimize txt These details include the initial values of the objective variable and the optimization variables the various realizations of the system being evaluated the array of viable systems or realizations population of individuals in genetic algorithm parlance by generation or as the optimization progresses and the final results A sample optimize txt file is illustrated in Figure 6 3 below P optimize_final txt Notepad File Edit Format View Help OBJECTIVES INITIAL VALUES waste_sys cost 220016600 000000 OBJECTIVE VARIABLES 2 Optimization Parameters maxInitialEvaluations 2000 maxPopulation 200 maxGenerations _ 30 optConvergenceLimit 1 000000000000000E 010 mutationfreg 1 0000000
60. ed for optimization The optimization free variables are determined from the variables of the component being optimized as well as other components flagged as its subcomponent Essentially subsystems are simply components that consist of several other components by virtue of the model equations its model equations consist of the declaration of other components If this component is a sub system that can be defined integrally and separately from other sub systems the optimization procedure will proceed faster by indicating that the component is a subsystem and interacts with other subsystems via a finite and few number of variables Then the subsystem is optimized in detail This procedure will proceed faster than calling an optimization command on each component and then calling an overall optimization command on the subsystem ILGO OPTIMIZATION GLOBAL OPTIMIZATION When a project has been defined into several subsystems each consisting of several components as illustrated in Figure 5 1 an optimization command may be called at the project level In this case iSCRIPT will perform a detailed optimization of each 95 subsystem and an overall optimization of the entire system project utilizing the sub system level coupling between the sub systems B 2 Procedures for Performing a Detailed Optimization in iSCRIPT To optimize a component the following information must be provided The objective variable from the list of componen
61. een of this program is shown below The underlying code modeling the components and subsystems depicted graphically would be the iSCRIPT platform Sse cuca D4 Graphics Controls _ Graphics_ x 1 Create components re Fala 2 Enter component variables alele EEE e CBE RERREEECEES fe Le eo LL Le m 3 Indicate coupling variables Exit Select Option From Icon Panel 4 Construct component model b Process for developing component model a System consisting of several components Figure 9 1 A graphical system building tool to complement iSCRIPT 13 Appendix A iSCRIPT Language Reference A 1 Variables and Expressions In stand alone scripts variables are declared as follows var1 var2 as type var1 var2 variable names satisfying the variable naming convention as declaration keyword type may take values logical short long real and double A 1 1 Types of Variables Logical variables Logical variables take on values of T or F In addition values of 0 or any real number may be assigned to logical variables A numeric value of 0 will be converted to F prior to assignment while other numbers will be converted to T Short variables Short variables have values in the range 32 768 to 32 767 They are also type INTEGER 2 in FORTRAN The syntax for the declaration al
62. f 0 001 L 1 f 0 002 L 2 f 0 003 L 3 f 0 004 L 4 f 0 06Re L 5 f 0 006 L 6 f 0 007 any other value of L i 1 f f 0 01 while i lt 2L Re Re 0 01 i 1 Input File eguations6g isc program rho v L mu Re f i as real Initial Values rho 1 05 v 17 0e 4 LSS 10 mu 8 5E 5 OANDABRWN 12 Models 13 14 Re Reynolds rho v L mu 15 16 mu 8 5E 5 17 call Computef Re f 20 Comment line here 22 L 5 24 end program 27 function Reynolds rho u L mu 28 rho u L mu Reynolds as real 29 Reynolds rho u L mu 30 u 12 2 31 end function 34 subroutine Computef Re f 35 Re f L as real 36 i as integer 38 L 5 40 if Re lt 2500 then Testing an if segment in a subroutine 41 f 64 0 Re 42 end 43 if Re gt 2500 then Testing a nested if segment in a subroutine 44 f 16 0 Re 0 4 45 end 46 if Re then testing a logical statement 47 if Re then testing a logical statement 48 if L 1 then 49 f 0 001 50 elseif L 2 then 51 f 0 002 52 elseif L 3 then 53 f 0 003 54 elseif L 4 then 55 f 0 004 56 elseif L 5 then 57 f 0 06 Re 0 4 58 elseif L 6 then 59 f 0 006 60 else 61 f 0 007 62 end if 63 end 64 end 66 L 4 ResetL 68 i 69 while i lt 2 L then testing a while loop segment 70 f f 0 01 71 Re Re 0 01 72 i i 1 73 end 74 end subroutine 113 Output F
63. ff_sup 2 4110405E 09 FF_fuselage 1 166486 Cdo_fuselage 3 8654767E 03 V 856 0000 M 0 8500000 Double Variables 5 rlong 0 000000000000000E 000 rlonger 0 000000000000000E 000 rnumber 0 000000000000000E 000 rnumb 0 000000000000000E 000 rexp 0 000000000000000E 000 117 Sample Problem 8 This script was derived from an old FORTRAN code to calculate friction factor The script uses expressions for Darcy Weisbach Colebrook or Churchill equations depending on Reynolds number calculated from the input The Colebrook equations require an iterative condition to obtain convergence based on a convergence criterion The output was compared to that of the equivalent FORTRAN program Model _ PUL u Re Laminar Flow 64 R lt 2100 J R Colbrook White Equation 472i 251 Jf 37D RAT Churchill Equation g 1 8 4 or By A pasties 37530 R R gt 4000 1 12 16 1 77K Saal Input File equations 13 isc The input file is not reproduced here but can be obtained directly from the SampleScript folder of the iSCRIPT installation Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 5 niter 0 iter 3 ilam 2 ierr 0 nitre 200 Real Variables 32 diam 1 000000000000000E 002 rey 659800 000000000 fric 3 794308979507086E 002 fric2 3 804064733943083E 002 fric3 9 69990906335253 1E 005
64. for Problem 5 5 Tower isc This is the routine for the cooling tower component subroutine Tower i Tower Q Condenser Q Tower T out Tower Q 3 7 Tower amp HR_sys w 1 2 Tower C 800 0 Tower A 0 6 end subroutine Figure 5 14 The tower component evaluation program for Problem 5 5 Pump isc This is the routine for the pump component subroutine Pump i Pump C 0 0005 HR sys w 3 0 end subroutine Figure 5 15 The pump component evaluation program for Problem 5 5 Condenser isc This is the routine for the Condenser component Subroutine Condenser i Condenser T_in Tower T_out Condenser C 270 Condenser T_in end subroutine Figure 5 16 The condenser component evaluation program for Problem 5 5 The code can be found in the subfolder SampleScripts HeatRejection Example5 3 of the iSCRIPT installation folder Follow the steps below to run this code 1 Open iSCRIPT Editor 2 Select File gt Open from the menu 49 gt The File Open dialog box appears 3 Navigate to the subfolder SampleScripts HeatRejection Example5 3 of the iSCRIPT installation folder 4 Open the file HeatRejection ipr 5 Select Tools gt Run Current Script Project File from the menu The result obtained is 31 846 19 for the total cost of the system 50 6 Optimization in iSCRIPT In this chapter we introduce the optimization procedure in isCRIPT Two examples will be given to
65. g the command mcc m PS_conv An executable is generated names PS_conv exe This executable was renamed to PS exe PH Run iSCRIPT Enter the project file ata ipr at the iSCript prompt View the results 5 Note that the model may be further optimized on the high level using the optimization procedures present in the iSCRIPT program 130 Sample Problem 18 Test of Input Output and Directory Management Commands This sample problem file illustrates the various input output and directory manipulation commands The file is equations24 isc and is listed below The model and output is the same as that in Sample Problem 1 Test of open close read write alt T P mu rho alt_l var2 var4 as real imyopen imyoout iout irun as integer var3 3 as real imyopen open input txt rt call write 6 File input file opened on unit imyopen imyoout open output txt wt call write 6 File opened output file on unit imyoout call write 6 Press any key to continue call read 6 call read imyopen g6 alt_1 var3 fscanf imyopen write Enter a value for var4 call read 6 var4 alt alt_1 0 3048 T 2 29013E 12 alt alt alt 8 60446E 08 alt alt 6 82246E 03 alt 3 18150E 02 P 101325 9 63714E 19 alt alt alt alt 1 18488E 13 alt alt alt P P 101325 5 52991E 9 alt alt 1 18225E 4 alt 1 mu 1 61988E 7 T T T 4 66143E 4 T T 4 7 25242E 1 T 4 20454 le 7 rho P 287 T irun
66. gument is optional default_value A default value for the variable all the variables for an array variable The type of this argument depends on type This argument is optional unit A string representing the engineering unit used in providing the variable values e g m s This argument is also optional and when provided is limited to 20 characters NOTE THERE IS NO NEED TO ENCLOSE CHARACTER ARGUMENTS IN QUOTES FOR THIS COMMAND A 9 3 Executing a Component A component may be executed using the following syntax Component_name execute Or Call Component_name execute Component_name _ Character 24 The component name as defined in 4 9 1 The execute routine must be a subroutine with the same name as Component_name and require no arguments 93 A 9 4 Using Global Variables in the In Built Global Component Global variables may be created only in the main program Global variables do not need to be created using the CreateVariable command although this command may be used as well Instead global variables may be created simply by prefixing a declaration with the Global keyword as in Section 2 1 Global Re Ma as real Global variable names must be unique among variable names but may coincide with a local variable name Reference to global variables is similar to that for all components as illustrated in the example below program global emCp Q1 as real localemCp Q r as real global emCp 209 4 r 4
67. h are S 2nr 2arh V ar h where r 5 and h 10 The equations for calculating the surface area S and volume V of a circular cylinder with 4 2 4 3 We will present a crash course in subroutines and functions in this section so that we can use them for the sample calculations being discussed Please refer to Appendix A 5 2 through A 5 5 for more details on subroutines and functions in iSCRIPT iSCRIPT supports the use of subroutines and functions In iSCRIPT a subroutine may be written as subroutine subroutinename arg1 arg2 argN variable declaration statements 32 end subroutine represents one or more lines of scripting language segments argl arg2 argN are the arguments to the subroutine variable declaration statements represent variable declaration statements for the arguments A subroutine may be called by call subroutinename arg1 arg2 argN where arg1 arg2 argN are the actual arguments to the subroutine Similar to a subroutine a function may be written as function functionname arg1 arg2 argN variable declaration statements end function represents one or more lines of scripting language segment argl arg2 argN are the arguments to the function variable declaration statements represent variable declaration statements for the arguments Functions may be called simply by using
68. haracters This variable must be a variable previously declared for this component using the CreateVariable command component2 The component to which the variable belongs a string limited to at most 24 characters This component must be the component to be optimized or a sub component of it This component must be a component previously declared with the CreateComponent command variable2 Name of the variable a string limited to at most 24 characters This variable must be a variable previously declared for this component using the CreateVariable command The above specification is interpreted as Component1 Variable1 Component2 Variable2 If the coupling is a function e g Heat_Ex1 Q 1 2 SecondaryHeat_Ex Q MCP T2 T1 An additional component variable may be created SecondaryHeat_Ex Q_Couple In the system model this variable may be set as SecondaryHeat_Ex Q_Couple 1 2 SecondaryHeat_Ex Q MCP T2 T1 This variable may then be coupled to Heat_Ex1 Q B 2 8 ILGO Optimize Command A system may be optimized using the following syntax System Optimize When the system optimize command is invoked iSCRIPT searches for every sub system defined in the project and optimizes each one Then an ILGO optimization is performed as described in Section 11 6 using the coupling variables 101 B 2 9 Performing Detailed Optimization at the System Level without ILGO A system level global optimization may also be
69. he optimization variables and the objective function the various realizations of the system being evaluated the array of viable systems or realizations population of individuals in genetic algorithm parlance by generation or as the optimization progresses and the final optimized results 15 3 Developing and Executing an iSCRIPT Project In this chapter we introduce the most basic 1iSCRIPT procedures for solving an arithmetic and algebraic problem evaluating the performance of an engineering system and optimizing an engineering system 3 1 Using iSCRIPT to Solve Basic Arithmetic and Algebraic Problems iSCRIPT has a built in functionality for arithmetic and algebraic operations This enables iSCRIPT to perform most mathematic calculations like some common symbolic computation software such as Matlab Maple Mathematica and Mathcad and programming languages such as FORTRAN and C In the current version of iSCRIPT a math problem is solved by wring a small script This is most conveniently done in the iSCRIPT Editor environment A GUI is being developed that will allow the user to directly evaluate the values of mathematic equations in a command window Open iSCRIPT Editor and create a new script file isc General code format 1 Declare the variables appear in the equations 2 Type the text form of the equations Write the script code in the file e Declare variables Typ
70. he start of the loop enddo or end may also be used instead of end do do statements may be nested as desired For Loops for ii expression expression2 end for The following rules govern the use of the for statement ii is a variable declared as in Section 4 1 ii may be a short long real or double variable expression1 and expression2 are valid expressions constructed as in Section 4 2 represents one or more lines of scripting language segments which may include other for statements and is referred to as the body of the loop ii is incremented by 1 and the body of the loop executed until expression1 is greater than expression2 The body is not executed at all if expression1 is greater than expression2 at the start of the loop endfor or end may also be used instead of end for for statements may be nested as desired 80 A 4 3 While Loops while expression 7 end The following rules govern the use of the while statement expression 1 is a valid expressions constructed as in Section 4 2 represents one or more lines of scripting language segments which may include other for statements and is referred to as the body of the loop The body of the loop is executed until expression1 evaluates to false or 0 The body is not executed at all if expression1 is false or evaluates to 0 at the start of the loop while statements may be nested as desired A 5 Subprogram and Function iSCRIPT allows t
71. he use of subprograms to introduce program structure and allow the organization of parts of the model For instance a Subroutine or function may be generated separately and called multiple times to perform a specific purpose Subprograms are useful in creating codes or models that are easily maintained and help to avoid rewriting whole segments of code that may be required more than one time Scripts may be created without any particular start or end program indicator In this case the entire script is assumed to be one program and subprograms can not be used To use Subroutines and subfunctions a start and end program indicator must separate the program and start and end subprogram indicators must also be used All scripting elements outside of these demarcators are ignored The exact syntax for structure limiters are described in this chapter A 5 1 Program Structure program programname variable declaration statements end program The following rules govern the use of program structure statements 81 A 5 2 programname is the name of the program and must be contrived according to variable naming conventions described in Section A 1 2 variable declaration statements represents several lines of variable declaration statements as described in Section A 1 1 represents one or more lines of scripting language segments and is referred to as the body of the program endprogram may also be used instead of end program There can
72. he various heat exchangers in the sub systems of aircraft Component tools are typically treated as a black box in the integration of models into the complete system Approximation Tools Response surfaces may be generated from measurements of the performance of a component as a function of selected decision variables and used as the model in the multi level optimization phase 1 1 3 Optimization The use of a detailed representation of a component is a critical factor in terms of computational resources The problems of interest are typically multi objective with objectives such as drag minimization or maximization of lift drag ratio in a mission segment gross take off weight minimization fuel consumption minimization minimization of acoustic noise during take off and landing cost minimization capital operating and environmental Although the capability for multi objective optimization is usually important a recombination of the various objectives into a single metric such as exergy may sometimes alleviate the multi objective requirement iSCRIPT optionally supports exergy based optimization of aircraft Optimization procedures include gradient formulation and genetic algorithms GA Gradient based methods work well for subsystems with continuous variables but are prone to local optima and divergence when the initial guess solutions are far from the true solutions Procedures based on genetic algorithms
73. iSCRIPT User s Manual TTC Technologies Inc P O Box 1527 Stony Brook New York 11790 June 2006 August 2006 December 2006 April 2007 July 2007 November 2007 2006 2007 Copyright TTC Technologies Inc Table of Contents Introduction 1 1 1 2 1 3 1 4 Introduction to Industrial Design Optimization Why We Developed iSCRIPT How We Want You to See iSCRIPT Conventions Used in this Manual Installation 2N Lids 2 3 2 4 2 3 2 6 System Requirements Required Software Installing the Required Software iSCRIPT Files Running Sample Problems Output Files Developing and Executing an iSCRIPT Project 3 1 3 2 3 3 Using iSCRIPT to Solve Basic Arithmetic and Algebraic Problems Using iSCRIPT for Performance Analysis of Engineering Systems Using iSCRIPT to Optimize an Engineering System Using iSCRIPT in Simple Problems 4 1 4 2 4 3 4 4 Getting Started A Simple First iSCRIPT Program Calculate 2 3 Another Simple Problem Aircraft Drag Calculation Another Arithmetic Problem Circular Cylinder Geometry Engineering Component Modeling in iSCRIPT 5 1 5 2 5 3 5 4 I 5 6 Component Modeling in iSCRIPT Creating Subsystems and Systems in iSCRIPT Solving a Problem Using Component Modeling or Decomposition Method Writing a Program in Several Script Files Example of a Program Developed in Several Script Files Example of a System with Several Components ii _
74. iable assignment part for the experimental data 3 This is the calculation of the coefficient of drag by using Em 4 1 4 This is the calculation of the drag force for the aricraft at 11 velocities between 0 mph and 200 mph end program Figure 4 5 Program outline for problem 4 2 Figure 4 5 shows the program outline for this problem You may type this iSCRIPT code yourself in isCRIPT Editor and replace the comment statements of steps 1 through 4 with the appropriate scripts Note In iSCRIPT any line starting with or is a comment line Any part of a line starting with or not within a string quote is also a comment Comments are not evaluated and are provided only for the convenience of the modeler to communicate details of the model or script to themselves or others While writing the code you may need to use the decision structure and loop structure iSCRIPT supports decision and loop statements of most popular programming languages For example the most common decision statement the if statement can be used in iSCRIPT as if expression then else end if represents one or more lines of scripting language segment If expression is true the first part will be executed Otherwise the second part will be executed The most common loop statement the do statement can be used in iSCRIPT as do ii expression expression2 end do
75. ibed in Section A 1 label words must not be declared e goto or go to may be used e represents one or more lines of scripting language segments within the same program or subprogram Example 1 using the continue keyword Re rho U L mu If Re lt 2500 then f 16 Re go to 2000 end if f 0 0064 Re 0 4 2000 continue Example 2 using a label with an expression Re rho U L mu 86 If Re lt 2500 then go to 2000 end if f 0 0064 Re 0 4 2000 f 16 Re A 7 Intrinsic Functions Below is a list of supported intrinsic functions Their arguments and characteristics are the same as their FORTRAN equivalents This list is constantly increasing Please check our website for an updated list at any time cos sin tan exp log log10 sqrt acos asin 10 atan 11 cosh 12 sinh 13 tanh 14 anint 15 aint 16 abs 17 real 18 dble 19 alog10 20 alog 21 sizeof 22 length 23 sum 24 avg 25 min 26 max OOO Or N gt A 8 Input Output iSCRIPT includes commands for input output to screen keyboard and files The commands are described in this section 87 A 8 1 Opening a File A file may be opened using the open command The syntax is as shown below call open unit filename permission_mode arg1 arg2 argN or unit open filename permission_mode arg1 arg2 argN The following rules govern the process of
76. ign the input values HR_Sys Execute end program Figure 5 12 The main program for Problem 5 5 In the main program the upper bound lower bound and default values of the component variables are set during the component variable declaration by using CreateComponent statement These values are summarized in Table 5 2 below System and Variable Upper Lower Default Renee Component Bound Bound Value Heat Rejection Coys 1 0E9 0 0 1 The upper bound and lower bound of System w 500 0 10 0 220 0 a design variable is selected to fit the Cinet 1 0E9 0 0 design constraints A 500 0 10 0 105 0 2 The default value can be assigned ne Qrower 1 0E9 0 0 during the variable declaration to tout 500 0 0 avoid extra input value assignments in Pump Cine 1 0E9 0 0 the body of the program Chee 1 0E9 0 0 3 The upper and lower bounds of O indana 10E9 0 0 unconstraint variables can be set to be Condenser bin 500 0 0 large or smaller enough so that the variables can never reach the constraints Table 5 2 The upper bound lower bound and default values provided for the component variables in Problem 5 5 48 HR_sys isc This is the main controlling routine for the Heat Rejection System subroutine HR_sys Condenser Q 0 14E 08 Tower Execute Purp Execute Condenser Execute HR_sys C Tower C Pump C Condenser C end subroutine Figure 5 13 The subsystem program
77. iles project file main program file HR_sys system file Tower component file Pump component file and Condenser component file and are shown in Figures 5 11 through 5 16 Heat Rejection ipr project main isc HR_sys isc Tower isc Pump isc Condenser ise Figure 5 11 The project file for Problem 5 5 47 main isc This is the main program for the heat rejection system program main 1 Create the subsystem Create the Tower component and its variables CreateComponent Tower the cooling tower CreateVariable Tower C double 0 0 1 0E9 0 0 0 0 CreateVariable Tower A double 0 0 500 0 10 0 105 0 m 2 CreateVariable Tower Q double 0 0 1 0E9 0 0 0 0 W CreateVariable Tower T_out double 0 0 500 0 0 0 0 C Create the Pump component and its variables CreateComponent Pump the pump CreateVariable Pump C double 0 0 1 0E9 0 0 0 0 Create the Condenser component and its variables CreateComponent Condenser the condenser CreateVariable Condenser C double 0 0 CreateVariable Condenser Q double 0 0 F CreateVariable Condenser T_in double 0 0 500 O 2 Create the systems Create the heat rejection system and its variables CreateComponent HR sys Heat_Rejection_ System CreateVariable HR_sys C double 0 0 1 0E9 0 0 0 0 CreateVariable HR sys w double 0 0 500 0 10 0 220 0 kg s 3 Assign the input values Tover 170 HR_sys w 200 3 Ass
78. illustrate system optimization 6 1 Optimization Based on Component Modeling iSCRIPT s built in optimization procedure is based on the decomposition of systems into components isCRIPT can optimize a single component or a large system containing several subsystems or components by using the integrated local global optimization ILGO technique In its most basic form every iSCRIPT optimization job must consist of one system and at least one subsystem consisting of at least one component By inference the system subsystem component hierarchy for the most basic job is then represented by one component which is also the subsystem which is the system To optimize a component or subsystem we must indicate the variable that should be maximized or minimized This variable is termed the objective variable and is indicated by using the AddObjective statement The details of this statement are as follows AddObjective component variable maxmin Note Segments enclosed in square brackets are optional and may be omitted component The component to be optimized a string limited to 24 characters This component must be a component previously declared with the CreateComponent command variable Name of the objective variable a string limited to 24 characters This variable must be a variable previously declared for this component using the CreateVariable command maxmin 0 or 1 Indicates whether thi
79. ipe component and its variable diameter D and cost CreateComponent Pipe CreateVariable Pipe D CreateVariable Pipe cost Create pump component and its variable gas flow rate w and cost CreateComponent Pump CreateVariable Pump w CreateVariable Pump cost Create gas exhaust system and its variable total cost CreateComponent Exhaust_sys CreateVariable Exhaust_sys totalcost 110 0 020 Pipe D Pump w Exhaust_sys Execute end program subroutine Pipe Pipe cost 160 Pipe D end subroutine subroutine Pump Pump cost 2 2E8 Pump w 2 end subroutine subroutine Exhaust_sys Pipe Execute Pump Execute Exhaust_sys totalcost Pipe cost Pump cost end subroutine Figure 3 2 The basic procedure for performance analysis of an engineering system in iSCRIPT 19 3 3 Using iSCRIPT to Optimize an Engineering System iSCRIPT s built in procedures can automatically optimize an engineering system that is formulated in component decomposition form as illustrated in Section 3 2 The way to optimize an engineering system assuming a single objective variable is similar to that for performance analysis except an objective variable must be indicated in addition to the requirements for system performance evaluation optimization variables must be indicated in addition to the requirements for system performance evaluation indicate the relations between the system components and the sys
80. isc i y x 5 b 5 z a c d e as integer b 12345 x 1 1 x 2 2 x 3 3 x 4 4 x 5 5 Do add up all the elements of x use this a sum b which is better than this z length x for i 1 length x y y x i end Joc avg x d min x e max x d min X e max X Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 9 i 6 y 15 x 5 1 2 5 b 5 1 2 5 Z 5 a 15 c 7 0 d 1 e 5 Real Variables 0 Double Variables 0 122 File Edit View Text Debug Breakpoints Web Window Help Dae teene Sms OH AE st b 1234 5 Do add up all the elements of x use this sum b which is better than this length x y 0 for i 1 length x Y ytx ij end c avg x d min x e max x Ln1 Col4 4l0 212 511 y x btx 2 sinty 4 2 33 4 6 6 7 8 9 2 l 1 1 2 1 1 1 2 4 1 0 2 1 2 511 y x btx 2 y x x 2 sin y 2 sin y 8 22 06 7 51 AM 123 Sample Problem 13 MATLAB Compatibility This script was used to illustrate compatibility with MATLAB matrix manipulation features in particular the contraction of matrix order The equivalent MATLAB file is matlabal0 m Model Input 4 2 3 1 File matlabal0 isc x 3 3 y 3 1 a 1 3 as integer x 4 6 6 y 0 a 2 1 3 bastell x 4 2 3 4 6 6 7 8 9 78 9 0 y 1 0 0
81. itialEvaluations 5000 Global maxGenerations 100 Global optconvergencelimit 1 0E 7 Global mwatationfregq n Fee l Global sampsizepervariable 500 Global maxILGOsteps 30 HR Sys Optimize end program Figure 6 4 The main program for Problem 6 2 Figure 6 4 is the complete code for the main program Since all of the other files are the same as those in Problem 5 3 component models are portable from solution to solution only the main program file is provided here Note the keywords Global maxPopulation Global maxInitialEvaluations Global maxGenerations Global optconvergencelimit Global mutationfreq sampsizepervariable and Global maxlLGOsteps are the parameters to control the performance speed and accuracy of the GA optimization calculation The default values are usually sufficient to solve most problems The default values are provided in the iSCRIPT Optimization Reference in Appendix B 2 11 However in the above program we have modified these defaults to obtain results faster Detailed information of these parameters can also be found in Appendix B 2 11 59 The scripts solving this problem can be found in the subfolder SampleScripts HeatRejection Example6 2 of the iSCRIPT installation folder Follow these steps to run the code 1 Open iSCRIPT Editor Select File gt Open from the menu gt The File Open dialog box appears 3 Navigate to the subfolder SampleScripts HeatRejection Example6 2 of
82. l cylinder These decision variables do not require input values The objective variable is the surface_area and the optimization variable is the radius The height is a dependent variable computed from the volume and radius In the main program the Cylinder component and its four component variables are declared first The input values are then assigned followed by the declaration of the objective and optimization variables The last part of the main program is the optimization calculation In the Cylinder component model the height of the cylinder is computed from the volume and the radius Then the value of the objective variable surface_area is calculated Figure 6 1 shows the program outline for this problem CircularCylinderOptimizeDemo isc Program main 1 Create the Cylinder component Create the four component varialbes of the Cylinder component Assign the input value for the volume Declare the objective variable surface_area Declare the optimization decision variable radius Start optimization calculation end program Subroutine Cylinder i 7 Calculate the height from volume and radius 8 Calculate the surface_area end subroutine Figure 6 1 Program outline for Problem 6 1 The complete code is shown in Figure 6 2 If you compare it with the code in Figure 6 1 the meaning of each statement is obvious 54 Note The genetic algorithm in iSCRIPT will be used for the optimization procedure Detail
83. l Values rnumber 4 Models rfactorial FACTORIAL mmumber end program Recursive calculation of the factorial of a number Function FACTORIAL n n m as integer FACTORIAL as integer ifn m 1 elseif n m 1 else rn n FACTORIAL 1 end if FACTORIAL rn return End Function FACTORIAL Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 2 rnumber 4 rfactorial 24 Real Variables 0 Double Variables 0 120 Sample Problem 10 This script was used to illustrate the use of the passing of literal valued arguments and expression arguments to functions In addition the syntax function calls within functions are illustrated The model is the same as in Sample Problem 6 Input File equations6m isc The input file is not reproduced here but can be obtained directly from the SampleScript folder of the isSCRIPT installation Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 0 Real Variables 7 rho 1 05000000000000 v 12 2000000000000 1 5 00000000000000 mu 8 500000000000001E 005 re 105 000000000000 f 4 932557310067764E 002 i 0 000000000000000E 000 Double Variables 0 121 Sample Problem 12 MATLAB Compatibility This script was used to illustrate compatibility with MATLAB matrix manipulation features The equivalent MATLAB file is matlabal m Input File matlabal
84. le projectl ipr The source files could also be individually entered at the command line Input File projectl ipr Output The output is similar to the Sample Problem 5 129 Sample Problem 17 Integration with a MATLAB program This solution consists of the following files PS_conv m ata ipr Main isc PS_component isc a MATLAB program that rates a low bypass turbofan aircraft engine The input to the program includes the altitude Mach number etc This program reads the input from a file PS_input txt The output from the program includes several variables including the thrust fuel consumption etc The output from the program is written to a file PS_output txt an iSCRIPT project file containing three files main isc ps_component isc ps_setting isc These files are described below a main program defining two components PS_setting and PS_Component PS_Setting simply sets the conditions for computing rating the aircraft engine an iSCRIPT component model file This model file includes commands to execute the MATLAB model An input file is created for the MATLAB program and the output from the MATLAB is read The output is further used to compute certain quantities including the total exergy destruction in the engine Procedures for Running this Solution 1 Run the MATLAB program in MATLAB to check the model of the aircraft engine 2 Compile the MATLAB program into an executable This step includes issuin
85. ling the component After the component subroutines type the keyword subroutine SystemName and end subroutine to create the system subprogram for the overall system In the body of the subroutine a evaluate each component of the system by typing ComponentName Execute in the system subprogram b if there are interactions between the components type the interaction equations c type the overall system performance evaluation equations Save the file in isSCRIPT Editor as an isc file Select Tools gt Run Current Script Project File from the menu to run the iSCRIPT code View the output on screen or in the file outputscript txt The details of these procedures are illustrated in sample problems in Chapter 5 18 Create the script file in iSCRIPT Editor Decompose the engineering system Create the main program Create components by using CreateComponent statement Create component variables by using CreateVariable statement Create system by using CreateComponent statement Create system variables by using CreateVariable statement Assign the input value for system or component variables Execute the system by using system_name Execute Simulate each component Create the subroutine for each component Type all the equations for the component Simulate the overall system Create the subroutine for the system Execute the components in order by using
86. llustrating specific parts of the scripting language syntax The sample problems their purpose and results are presented below They are also included in SampleScripts folder of the iSCRIPT installation You may use these sample problems to gain familiarity with iSCRIPT syntax or copy any portion of the files for use in your own script Sample Problem 1 This problem illustrates the use the declaration of variables the reading and interpretation of expressions and the results Notice that the program is free flow and without a start or end program indicator iSCRIPT is able to execute free flow scripts without any particular program structure or subroutines Model T 2 2901x10 alt 8 60446 x 10 alt 6 82246 x 10 alt 31 815 9 63714 x 107 alt 1 18488 x 10 alt 5 52991 x10 alt 1 18225 x 10 alt 1101325 P f 1 61988 x 10 7 4 66143 x10 T 7 25242 x 1077 4 10454 107 me x P 38T where alt is the altitude in m T the temperature in K P the pressure in Pa and u the viscosity in Ns m and p the density in kg m The input value of alt is supplied in ft alt_1 and has to be initially converted to m in the script below Input File equations2 isc 1 at T P mu rho alt_1 2 alt 1 3000 3 alt alt_1 0 3048 4 T 2 29013E 12 alt alt alt 8 60446E 08 alt alt 6 82246E 03 alt 3 18150E 02 5 P 101325 9 63714E 19 alt alt alt alt 1 18488E 13 alt alt alt 6 P P 101325 5 52991E 9 alt alt 1 1822
87. lobal minimum Figure P19 1 Plot of the Rastrigin function The file is rastrigin isc and is listed below This is the system program This program defines all components subsystems and systems program main global pi as double pi 4 atan 1 0 133 1 Create the components 2 Create the subsystems One component One subsystem one system Create the entire system component and its variables CreateComponent Rastrigin Models_entire_system CreateVariable Rastrigin y double 0 0 1 0E14 0 0 0 0 Create Variable Rastrigin x1 double 0 0 10 0 10 0 0 5 kg s Create Variable Rastrigin x2 double 0 0 10 0 10 0 0 5 m2 AddObjective Rastrigin y 0 AddVarObjective Rastrigin x1 AddVarObjective Rastrigin x2 3 Evaluate the system at the initial conditions then optimize Rastrigin Execute Global maxPopulation 4000 Global maxInitialEvaluations 10000 Global maxGenerations 200 Global optconvergencelimit 1 0E 7 Global mutationfreq 0 2 Global sampsizepervariable 8000 Try Global maxPopulation 100 Global maxInitialEvaluations 500 Global maxGenerations 100 Global optconvergencelimit 1 0E 7 Global mutationfreq 0 2 Global sampsizepervariable 200 Try Global maxPopulation 100 Global maxInitialEvaluations 400 Global maxGenerations 100 Global optconvergencelimit 1 0E 7 Global mutationfreq 0 2 Global sampsizepervariable 100 Try Global maxPopulation
88. ly a sub system to which another component is contained within a string limited to 24 characters This component must be the component to be optimized or a sub component of it This component must be a component previously declared with the CreateComponent command sub component The component to which belongs to the sub system a string limited to 24 characters This component must be the component to be optimized or a sub component of it This component must be a component previously declared with the CreateComponent command NOTE THERE IS NO NEED TO ENCLOSE CHARACTER ARGUMENTS IN QUOTES FOR THIS COMMAND Example CreateComponent Heat_Ex1 CreateVariable Heat_Ex1 Tin CreateVariable Heat_Ex1 Tout CreateComponent Heat_Ex2 CreateVariable Heat_Ex2 Tin CreateVariable Heat_Ex2 Tout CreateComponent ECS CreateVariable ECS Weight CreateVariable ECS Drag AddSubComponent ECS Heat_Ex1 AddSubComponent ECS Heat_Ex2 The indication of a component as belonging to another is only used when an optimization command is called When the optimization command is called for a component iSCRIPT will search for optimization variables from the component and all components belonging to it to perform an overall detailed optimization of the component B 2 4 Component Optimize Command 99 A component may be optimized using the following syntax Component optimize Component A component name a string limited to
89. m represents one or more lines of scripting language segments and is referred to as the body of the program The first few lines of the program should consist of the variable declaration In iSCRIPT all of the local variables with the data type of variable should be declared before any executable statements The format of data declaration in iSCRIPT is varl var2 as type var1 var2 variable names satisfying the variable naming convention as declaration keyword type may take the values logical short long real and double see Appendix A 1 1 After the variable declaration the input data should be assigned to the variables More details about variable names variable types and permissible expressions can be obtained in the iSCRIPT Language Reference in Appendix A With the experimental data assigned the drag coefficient can be calculated by using Equation 4 1 Let us calculate the drag force on the aircraft for 11 velocities evenly distributed between 0 and 200 mph The corresponding drag force results and the velocities can be stored in two different arrays Note that iSCRIPT supports array operations and most of the intrinsic functions directly support vector and matrix operations More information on working with arrays is provided in the iSCRIPT Language Reference in Appendix A 28 AircrattDragDemo isc Program main 1 This is the variable declaration part 2 This is the var
90. mance analysis of the component An iSCRIPT file may also contain several subprograms and or functions 2 4 2 Main Program When an iSCRIPT solution includes at least one subroutine or function it is considered as having a program structure Every iSCRIPT solution with a program structure must contain a main program There must be only one main program in a solution When a solution has a program structure the execution of an iSCRIPT code starts from the main program When 11 iSCRIPT is used in a component modeling form the main program usually consists of two parts The first part is the component creation and component variable declaration and the second part contains the system evaluation and or optimization commands When a solution has no program structure and necessarily comprises of a single file the execution simply starts from the iSCRIPT file at the first line 2 4 3 Project Files ipr When a solution developed in iSCRIPT includes more than one iSCRIPT file an iSCRIPT project file ipr is needed to aggregate all the files The project file obeys the following rules 1 The file content must start with a unique keyword project on the first line 2 Both the name and directory of the script files should be included inside the project file When the script files and the project file are in the same folder the directory information of the script files may be omitted 3 At least one of the script files must
91. n square brackets are optional and may be omitted component The component to which the variable belongs a string limited to 24 characters Two component variables may not have the same name Component variable names obey the same formation rules as those for variables name A name for the component a string limited to 24 characters Two components may not have the same name Component names obey the same formation rule as those for variables type A string accepting values such as integer real double A complete list of variable types can be found in Section A 1 This argument is optional When not provided component variables are assumed to be double values dimension Variable dimension for an array variable integer For example a 2D matrix will have a dimension of 2 This argument is optional for scalar variables dimension 0 is default 92 size Variable size for an array variable This argument accepts an integer array with a limit of five integers For example a 3 x 3 matrix will have a size of 3 3 This input must be enclosed in brackets This input is required when dimension gt 0 upper_bound An upper bound for the variable all the variables for an array variable The type of this argument depends on type This argument is optional lower_bound A lower bound for the variable all the variables for an array variable The type of this argument depends on type This ar
92. nalytical results are r 3 je 5 0308 and m S 3V 27 477 0617 respectively Although the analytic results are easily obtained note the speed of the genetic GA optimization procedure used by iSCRIPT in calculating the results and the accuracy of the results Conventional wisdom would claim that GA procedures are robust but take significantly longer to converge or compute the optimum of continuous functions compared with gradient based methods However the unique GA procedure in iSCRIPT has been tuned to perform exceptionally well even for continuous functions while retaining the robustness expected of a GA procedure 6 3 Optimization of a System with Multiple Components Let us now consider a problem that contains more than one component The heat rejection system design problem in Problem 5 3 is used Problem 6 2 Minimize the Cost of the Heat Rejection System in Problem 5 3 56 The goal of the problem is to find the optimum value of A and w that will minimize the total cost C The objective variable of the problem is the total cost C of the Heat Rejection System There are two optimization variables flow rate of water w of the Heat Rejection System and the area A of the Cooling Tower component The objective function of this problem may be expressed as C f w A 6 2 The goal is to seek the optimum values of w and A that will minimize Csys Note that in this problem we have two
93. ns Otherwise this value should be set at a large number and the convergence limit discussed next used to terminate the improvement runs optconvergencelimit the improvement runs are terminated after the individuals in the top half of the population are no different from the previous generations by this value using the L2 norm mutationfreq this parameter sets the frequency of mutation per variable The parameter can be effectively used to control the procedure For instance if a specific problem is noted to be very susceptible to local optima or has a very narrow optimum window a higher value of the mutationfreg combined with more generation runs will ensure that the true optima are found Otherwise note 104 that using high values of mutationfreq will only make it take longer to settle on the optimum value maxilgosteps Similar to maxgenerations this parameter determines the number of ILGO improvement steps However this value is usually small considering that it nests within it several optimization runs within it A graph of the objective function over the ILGO steps is a good indicator of whether convergence has been reached Note that the parameters must be altered according to any global variable in iSCRIPT For example the maximum population size can be limited as follows Global maxpopulation 70 105 Appendix C MORE EXAMPLES OF iSCRIPT SYNTAX Additional sample problems were developed and used in i
94. nt l by Component l Execute Type the interaction equations between component l and component Z Execute component 2 by Component2 Execute Type the interaction equations between component N 1 and component N Execute component N by ComponentN Execute Type the overall system performance equations subroutine Example Find the optimum values of D and w to minimize the total cost of gas exhaust system Assume 90 lt D lt 150 and 0 015 lt w lt 0 055 program main Create pipe component and its variable diameter D and cost CreateComponent Pipe CreateVariable Pipe D double 0 0 150 90 110 CreateVariable Pipe cost double 0 0 10000000 0 0 Create pump component and its variable gas flow rate w and cost CreateComponent Pump CreateVariable Pump w double 0 0 0 055 0 015 0 020 CreateVariasble Pump cost double 0 0 10000000 0 0 Create gas exhaust system and its variable total cost CreateComponent Exhaust_sys CreateVarisble Exhaust sys totalcost double 0 0 10000000 0 0 Declare the objective variable AddObjective Exhaust sys totalcost 0 Declare the optimization decision variable AddVar0bjective Pipe D AddVar0bjective Pump w Indicate Pipe and Pump as subsystems of Exhaust_sys AddSubC omponent Exhaust _sys Pipe AddSubComponent Exhaust sys Pump Start optimization calculation Exhaust_sys Optimize end program subroutine Pipe Pipe cost 160 Pipe D end subrou
95. o other variables but must include the size and dimension of the arrays in brackets after the array names Array sizes must be integer values Examples include matrixA 3 3 as real array1 5 2 inumber as integer A 2 1 Referencing Array Elements Array elements may be referenced using numbers variables or expressions The type of the number variable or expression will be converted to integer at runtime Examples include matrixA 1 1 12 2 matrixA 1 2 Reynolds_No matrixA 1 3 rho U L mu matrixA j sin t r 2 k cos omega t Arrays can be included directly in expressions and array arithmetic performed A 2 2 Assigning Values to Arrays Literal values can be directly assigned to array variables For instance the segment below generates a 3 x 3 matrix x x 3 3 as integer x 4 2 3 4 6 6 7 8 9 A 2 3 Matrix Arithmetic Matrix arithmetic such as multiplication divisions additions etc may be performed directly An example is shown below i x 3 3 y 3 3 b 3 3 as integer 2 3 3 a 3 3 as real x 4 2 3 4 6 6 7 8 9 78 y 2 1 1 1 2 1 1 1 2 b 410 212 51 1 Z y x b x 2 y x a X 2 sin y A 3 Decision Structure iSCRIPT uses the if elseif else endif statement to implement the conditional execution of segments of a program The syntax is as shown below if expression then elseif expression2 then elseif expression3 then else end if Th
96. o those structures In iSCRIPT the structures can be created as components and the properties of the component are referred to as component variables However unlike normal structures all components automatically support the Component Execute method The syntax to define an object and the properties or variables of the objects is described below Note that a global component exists for every project as described in Section A 9 3 The global component has no execute file A 9 1 Defining a Component An object may be defined using the following syntax CreateComponent name description Note Segments enclosed in square brackets are optional and may be omitted name A name for the component a string limited to 24 characters Two components may not have the same name Component names obey the same formation rules as those for variables description A description for the component a string limited to 50 characters Optional 91 NOTE THERE IS NO NEED TO ENCLOSE CHARACTER ARGUMENTS IN QUOTES FOR THIS COMMAND Example program main T_in T_out as real CreateComponent Heat_Ex1 CreateVariable Heat_Ex1 Tin CreateVariable Heat_Ex1 Tout T_in 286 16 Heat_Ex1 Tin T_in A 9 2 Defining a Component Variable A component variable may be defined using the following syntax CreateVariable component name type dimension size upper_bound lower_bound default_value unit Note Segments enclosed i
97. odeling each subsystem in a decomposed fashion It is assumed that the propulsion subsystem PS is modeled in a separate executable file that receives parameter input representing variables such as the operating point Mach number altitude and amount of bleed air extracted for the environmental control subsystem ECS The iSCRIPT model integrates the PS program into the complete aircraft model by writing the parameter input into a file in the required format of the PS model running the PS program and retrieving the output from the output file in the program s output format 70 Figure 7 2 An aircraft model with the PS subsystem modeled in a software or executable outside of iSCRIPT The solution consists of the following files in the SampleScripts PS_inMATLAB subfolder of the iSCRIPT installation folder PS_conv m ata ipr Main isc PS_component isc a MATLAB program that rates a low bypass turbofan aircraft engine The input to the program includes the altitude Mach number etc This program reads the input from a file PS_input txt The output from the program includes several variables including the thrust fuel consumption etc The output from the program is written to a file PS_output txt an iSCRIPT project file containing three files main isc ps_component isc and ps_setting isc These files are described below a main program defining two components PS_setting and PS_Component PS_Setting simply sets the condi
98. omposition may not be apparent On the other hand a realistic industrial design optimization problem could contain many components and a large number of variables A system may contain several subsystems and the coupling between these subsystems may be complicated iSCRIPT s decomposition based modeling approach and ILGO optimization algorithm provide a way to handle these kinds of complicated system design and optimization problems 61 7 Running iSCRIPT in Parallel iSCRIPT can execute programs automatically in a parallel environment without the user actually parallelizing their codes However currently only the optimization portion of a program actually executes in parallel Optimization commands are computationally intensive for fairly sized system models As a result executing them in parallel provides significant reduction in the time to obtain solutions For instance a program named ADVISOR implemented in MATLAB which simulates in detail the model for an automobile drive train system is reported to take about 25 seconds to run on a fairly sized PC Optimizing such a program which is doubtlessly a mixed integer non linear or MINL problem may require up to 40 000 evaluations of the model This would result in a total time of about 11 days to obtain results In a parallel environment using 12 processors results may be obtained in as quickly as 24 hours Any iSCRIPT program that can run on a single processor computer can execute
99. onent to which the variable belongs a string limited to 24 characters Two component variables may not have the same name Component variable names obey the same formation rule as those for variables variable_name A name for the variable a string limited to 24 characters Two components may not have the same name Component names obey the same formation rule as those for variables 99 66 type A string accepting values such as integer real double A complete list of variable types can be found in Appendix A 1 This argument is optional When not provided component variables are assumed to be double values dimension Variable dimension for an array variable integer For example a 2D matrix will have a dimension of 2 This argument is optional for scalar variables dimension 0 is default size Size for an array variable This argument accepts a group of integers in a bracket separated by a semicolon with a limit of five integers For example a 3 x 3 matrix will have a size of 3 3 This input is required when dimension gt 0 upper_bound An upper bound for the variable all the variables for an array variable This argument is optional lower_bound _ A lower bound for the variable all the variables for an array variable This argument is optional default_value A default value for the variable all the variables for an array variable This argument is optional
100. ously declared with the CreateComponent command variable Name of the variable a string limited to 24 characters This variable must be a variable previously declared for this component using the CreateVariable command delta This variable further narrows the optimization search space for this domain to MAX L x A MIN M x A from L M where L and M are the lower and upper bounds for variable xi as defined in CreateComponentVariable x is the current values of the variable and A is delta It is useful to further reduce the search space after a prior optimization step or after an initial computation based on initial conditions This variable is also used internally to narrow the optimization search space in the ILGO procedure discussed later following the first ILGO step The type of the argument depends on the type of the variable NOTE THERE IS NO NEED TO ENCLOSE CHARACTER ARGUMENTS IN QUOTES FOR THIS COMMAND Example CreateComponent Heat_Ex1 CreateVariable Heat_Ex1 Tin CreateVariable Heat_Ex1 Tout CreateVariable Heat_Ex1 Length CreateVariable Heat_Ex1 Width CreateVariable Heat_Ex1 Weight AddObjective Heat_Ex1 Weight 0 AddVarObjective Heat_Ex1 Length AddVarObjective Heat_Ex1 Width B 2 3 Indicating Component Relationships A component may be flagged as contained within another component as follows AddSubComponent component sub component 98 component A component subsequent
101. outputvar Computef Re f Re f L outputl output2 as real i as integer outputvar as integer outputl 17 0 output2 21 0 L 5 if Re lt 2500 testing a logical statement f 64 0 Re end if Re gt 2500 then Testing a nested if in a subroutine 4 f 16 0 Re A 0 4 en if Re testing a logical statement if Re then testing a logical statement if Q 1 then f 0 001 elseif L 2 f 0 002 elseif L 3 then f 0 003 elseif Q 4 then f 0 004 elseif Q 5 then f 0 06 Re A 0 4 elseif 0 6 then f 0 006 else f 0 007 end if end end L 4 Reset L outputvar Re L outputl i l while G lt 2 L then testing a while loop segment f f OQ Re Re 0 01 is sj 1 if G 5 then break end if end function 127 Output FINAL VALUES OF VARIABLES Logical Variables 0 Integer2 Variables 0 Integer Variables 1 initest3 5712 Real Variables 9 rho 1 050000 v 1 7000000E 03 L 5 000000 mu 8 5000000E 05 Re 84 00000 f 0 0000000E 00 i 0 0000000E 00 rtest1 17 00000 rtest2 84 00000 Double Variables 0 128 Sample Problem 16 Multi Source File projects This script was used to illustrate the multi source file project capabilities of iSCRIPT The model is the similar to that in Sample Problem 6 The source files include Projectla isc Project1b isc Projectlc isc Projectld isc All source files were listed in a single project fi
102. ow 23 gt iSCRIPT file operations e To create a new code file you can click the new blank document D icon or select File gt New from the menu e To open an existing code file you can click the open file icon or select File gt Open from the menu to open a file opening dialog box Change the current file directory and select the file you want to open e To save a code file you can click the save file icon w or select File gt Save from the menu gt Operations for running iSCRIPT e ToruniSCRIPT you can select Tools gt Run Current Script Project FIle from the menu or use the hot key F8 e To analyze a code you can select Tools gt Compile Analyze Current Script Project File from the menu or use the hot key F7 Note The JSCRIPT Analyzer command outputs detailed information related to the results of evaluating every line of an iSCRIPT code for debugging purposes while JSCRIPT command does not provide detailed output For this reason JSCRIPT command executes faster than the Analyzer 4 2 A Simple iSCRIPT Program Calculate 2 3 In this section we will introduce the steps required to develop a script for an arithmetic and algebraic problem We use the simple arithmetic problem 2 3 as our first example We will write a short iSCRIPT program which we call main to solve this problem The procedure follows the general procedure outlined for arithme
103. p iscript_mp exe program ipr In the above syntax is the required number of processors program isc is any iSCRIPT program file program ipr is any iSCRIPT project file If the first syntax is used the iSCRIPT parallel program will initially interactively query the user for the script or project that the user wishes to execute in parallel 7 3 A Sample Optimization Problem Run in Parallel Any iSCRIPT program that can run on a single processor computer can execute in parallel in a multi processor environment This example illustrates the use of iSCRIPT in optimizing a problem in parallel This sample problem file contains a main program and a subroutine that evaluates a model The model is the Rastrigin equation shown in Equation 6 1 below f x 20 x x5 10 cos 27x cos 27x 6 1 Global minimum Figure 6 1 Plot of the Rastrigin function 63 This function has several local minima making it difficult for a gradient based procedure to capture the actual minimum without the benefit of a good starting or guess value The actual minimum value is 0 and occurs at the values of x1 x2 0 0 Figure 6 2 below shows the iSCRIPT program for finding the optimum of the Rastrigin function rastrigin isc This is the system program This program defines all components subsystems and systems program main global pi as double gpi 4 atan 1 0 1 Create the components 2 Create the components This sy
104. p isc Qexamples 2 Tower isc condenser isc B ction ipr HR_sys isc A main isc File name HeatRejection ipr Files of type All Files gt ASCII DOS READ REC COL OVR e The project file will open as shown below 13 iSCRIPT Editor C iScript SampleScripts HeatRejection HeatRejection ipr C File Edit Search View Document Project Tools Macros Window Help Desig Fe ak 4B mAn nAAL y 6 Text HeatRejection ipr Condenser isc wf Ln1 Ch1 ASCII DOS READ REC COL OVR Ose er A SL BE Preferences 6 Text HeatRejectionipr Evaluate ae Ctrl Enter MS DOS Shel F10 View in Browser Akt 8 w Load User Tools Tover isc Conf User Tools Pump isc Condenser isc Compile Analyze Current Script Project File F7 Run Current Script Project File F8 ay Execute or run the current iSCRIPT file or project Run File Ln1 Ch1 ASCII DOS AD REC COL OVR e The following screen will appear containing the output of the program MM sysdes Graphic1 ile argum 2 6 Output Files 14 The results of a calculation are stored in the file outputscript txt in the same folder of the project file which contains the computation results of all component variables You can also add your own I O using iSCRIPT I O commands For projects involving optimization additional information can be found in the file optimize txt The information includes the initial values of t
105. pr for Problem 5 2 Cylinder isc program main radius height surface_area volume as real pi as real radius 5 height 10 surface_area get area radius height call get volume volume radius height end program Figure 5 5 Script file Cylinder isc for Problem 5 2 Area isc function get _area r h get area r h as real pi as real pi 3 1415926535 get_area 2 pi r 2 2 pi r h end function Figure 5 6 Script file Area isc for Problem 5 2 Volume isc subroutine get_volume v r h V r h as real pi as real pi 3 141592 6535 pi tr 2 h end subroutine Figure 5 7 Script file Volume isc for Problem 5 2 The code can be found in the subfolder SampleScripts CircularCylinder Example5 2 of the iSCRIPT installation folder Follow the procedures below to run this code Open iSCRIPT Editor 43 Select File gt Open from the menu The File Open dialog box appears Navigate to the subfolder SampleScripts CircularCylinder Example5 2 of the iSCRIPT installation folder Open the file Cylinder ipr Select Tools gt Run Current Script Project File from the menu If the program is run same result will be obtained as in Problem 4 3 5 6 An Example of a System with Several Components We will use a heat rejection system to illustrate performance analysis of a system with several components Problem 5 3 Calculating the Cost of a Heat Rejection System
106. r Optimize end program subroutine Cylinder pi as real pi 3 141592654 Cylinder height Cylinder volume pi Cylinder radius 2 Cylinder surface area 2 pi Cylinder radius 2 2 pi Cylinder radius Cylinder height end subroutine Figure 6 2 The complete code for Problem 6 1 55 The code can be found in the subfolder SampleScripts CircularCylinderExample6 1 of the iSCRIPT installation folder Follow these steps to run this code 1 Open iSCRIPT Editor Select File gt Open from the menu The File Open dialog box appears 3 Navigate to the subfolder SampleScripts CircularCylinder Example6 1 of the iSCRIPT installation folder 4 Open the file CircularCylinder isc 5 Select Tools gt Run Current Script Project File from the menu The code returns the values of the minimized surface area and the corresponding cylinder radius as 477 06172 and 5 030 respectively Note that due to the fact that a genetic algorithm is used the results may change slightly with each execution Also note that default values of the optimization parameters in iSCRIPT have been used which is why no parameters are specified These defaults are usually acceptable for most problems Details of the parameters are given in Appendix B 2 11 The analytic results can be derived by differentiating the following equation and setting the value to zero to derive the optimum 2 S 2a r 8 2m r 4 6 1 7Yr For this problem the a
107. r isc program main radius height surface area volume as real pi as real radius 5 height 10 WOODIAmM fone surface_area get_area radius height call get_volure volume radius height end program function get_area r h get area r h as real pi as real pi 3 1415926535 get_area 2 pitr 2 2 pi r h end function subroutine get_volume v r h v r h as real pi as real pi 3 142415926535 v pitr 2 h end subroutine Figure 4 8 iSCRIPT program for evaluating the cylinder geometry This code can be found in the subfolder SampleScripts CircularCylinder Example4 3 of the iSCRIPT installation folder Follow these steps to run the code 34 Step 1 Open iSCRIPT Editor Step 2 Select File gt Open from the menu gt The File Open dialog box appears Step 3 Navigate to the subfolder SampleScripts CircularCylinder Example4 3 of the iSCRIPT installation folder Step 4 Open the file CircularCylinder isc Step 5 Select Tools gt Run Current Script Project File from the menu The results for the surface area and volume are 471 2389 and 785 3982 respectively 35 5 Engineering Component Modeling in iSCRIPT In this chapter we present the component modeling technique in iSCRIPT Examples are given to show the procedure for creating and executing engineering component modeling mode in iSCRIPT The procedure for developing an iSCRIPT solution in several script files i
108. rcraft drag calculation The problem is described below Problem 4 2 Aircraft Drag Calculation Relative cyrust Wind ee Direction of Flight Weight Figure 4 4 Aerodynamics Forces on an Aircraft In general the aircraft drag can be computed by pV A 2 drag C 4 1 where C drag coefficient which is usually determined experimentally in a wind tunnel p air density V velocity of aircraft A reference area the surface area over which the air flows Although the drag coefficient is not a constant it can be assumed to be at low speeds less than 200 mph Suppose the following data were measured in a wind tunnel drag 20 000 N P 1x10 kg m V 100 mph A 1m Calculate the drag coefficient then use the computed drag coefficient to predict how much drag force will be exerted on the aircraft at velocities from 0 to 200 mph 27 To solve this problem in iSCRIPT we need to write a program that computes the drag coefficient C4 from the wind tunnel data The drag coefficient is then used to calculate the drag force on the aircraft at a range of velocities between 0 and 200 mph As for the math problem in Section 4 2 the script may be created with a program structure In iSCRIPT the program structure starts with a keyword program programname and ends with a keyword end program The format of the program is program programname end progra
109. rs 103 There are five parameters in iSCRIPT that control the genetic algorithm used for optimization in iSCRIPT They are implemented internally as global variables with default settings their values may be reset and altered in any model file They are Parameter Default Value maxinitialevaluations 1000 maxpopulation 500 maxgenerations 8 optconvergencelimit 0 001 mutationfreq 0 01 maxilgosteps 5 maxinitialevaluations this parameter limits the initial search space size Otherwise the algorithm attempts to sample each variable at 10 points in its search space For a 10 variable problem the sample size is potentially approaching a fraction of the number C o The parameter should be set to a lower number since other combination and mutation operators used in subsequent generations reduces the need to sample excessively for the initial population maxpopulation this parameter limits the overall population size Otherwise the algorithm attempts to set a limit of 20 times the number of variables The population size slows down the genetic algorithm procedure and this parameter can be used to control the population size effectively without compromising the ability of the process to obtain the true optima maxgenerations this parameter the number of improvement generations to run This parameter is intended to be used if the user wishes to run the algorithm in several cycles effectively restarting a new cycle after maxgeneratio
110. s subsystems and systems in iSCRIPT 5 3 Solving a Problem Using Component Modeling or Decomposition Method We still use the circular cylinder surface area and volume calculation problem of Section 4 4 to illustrate the procedure We will now solve that same problem using the component method Problem 5 1 Solve Problem 4 4 Using the Component Technique in iSCRIPT For this problem we define a single component Cylinder which has four component variables radius height surface_area and volume The detailed procedure to solve the problem follows the general procedure in Section 3 2 The steps are as follows Step 1 Open iSCRIPT Editor Step 2 Create a new file and save it as an isc file Step 3 Type the words program main and end program to create a main program In the body of the main program a declare the Cylinder component by using CreateComponent statement 39 b declare the four component variables radius height surface_area volume of the Cylinder system using CreateVariable statement c assign the input values for Cylinder radius and Cylinder height d type Cylinder Execute to evaluate the component Step 4 After the main program is written type the words subroutine Cylinder and end subroutine to create the system subroutines for the Cylinder system In the body of the subroutine a type in the equations to calculate the surface area and volume Step
111. s also presented 5 1 Component Modeling in iSCRIPT iSCRIPT has features that allow you to define components and the variables associated with the components These variables will have an engineering context including limits constraints as well as engineering units where appropriate To complete the modeling of a component a component subprogram must be written that contains the equations to model the component The model equations for a component can be evaluated with the statement Component Execute In iSCRIPT a component is declared by using a CreateComponent statement All components of a system must be declared in the main program CreateComponent component_name description Note Segments enclosed in square brackets are optional and may be omitted component_name A name for the component a string limited to 24 characters Two components may not have the same name Component names obey the same formation rule as those for variables description A description for the component a string limited to 50 characters Optional After the component is declared variables can be attached to this component by using a Create Variable statement CreateVariable component_name variable_name type dimension size upper_bound lower_bounda default_value unif Note Segments enclosed in square brackets are optional and may be 36 omitted component_name The comp
112. s as components They include variables at the subsystem level similar to component variables The subsystem variables are variables at the subsystem level that cannot be isolated within any of the components For instance in an airframe subsystem consisting of components including say a fuselage wind tail and ailerons the total surface area as a variable would be a subsystem variable On the other hand the wing span sweep angle and thickness would be the wing component s variables and the fuselage length and diameter would be the fuselage component s variables Subsystems are also declared using the CreateComponent command and subsystem variables are also declared using the CreateVariable command The way subsystems differ from components is that they have other components associated within them The association is indicated using the AddSubComponent command However this formal association is only necessary for optimization tasks Chapter 6 and Problem 6 2 of Section 6 3 The mere evaluation of the subsystem model consisting of the evaluation of its component models is sufficient for the performance analysis of the subsystem see Figure 5 12 of Section 5 6 Systems do not need to be declared in iSCRIPT By default each iSCRIPT project is considered to be a system All subsystems and their components declared in a specific project are therefore assumed to belong to the system Problem 5 3 illustrates the relationship between component
113. s by using AddObjective statement Declare the optimization variables by using Add VarObjective statement Indicate the subsystems by using AddSubComponent statement Optimize the by system using System Optimize Simulate each of component Create the subroutine for each component Type all the equations for the component Simulate the overall system Create the subroutine for the system Execute the components in order by using component_name Execute statement Type the interaction equations between components Type the overall system performance equations Run the program by selecting Tools gt Run Current Script Project File program main Declare N components of system by using CreateComponent Declare the component variables for each component by using CreateVariable Declare the overall system by using CreateComponent Declare the overall system variables by using CreateVariable Assign the input values for the system or component variables Declare the objective variable by using Add0bjective Declare the optimization variables by using AddVarObjective Indicate the subsystems by using AddSubComponent th th HR TR HR HR OTR Oe Optimize the overall system by using System Optimize subroutine component Type all the equations for the component 1 end subroutine subroutine componentN i Type all the equations for the component N Execute compone
114. s is a minimization or maximization objective Use 0 to minimize and 1 to obtain a maximum This argument is optional The default value is 0 51 After the objective variable has been indicated we also need to indicate which variables are free for optimization These are the variables whose values can be varied in order to obtain the optimum value of the objective function The variables are referred to as optimization or decision variables and are indicated by the AddVarObjective statement The details of this statement are as follows component variable delta AddVarObjective component variable delta Note Segments enclosed in square brackets are optional and may be omitted The component to which the variable belongs a string limited to 24 characters This component must be the component that will be optimized or a subcomponent of it see Section B 2 6 This component must be a component previously declared with the CreateComponent command Name of the variable a string limited to 24 characters This variable must be a variable previously declared for this component using the CreateVariable command This parameter further narrows the optimization search space for the optimization variable If the optimization variable x were defined by CreateVariable to have the lower and upper bounds L and M respectively then x L M j and the size of the search space is M L The parameter A
115. s of the genetic algorithm are provided in Appendix B 2 10 Since the optimization includes a search process with many iteration steps both the objective variable and optimization variable must be set with a reasonable upper bound and lower bound constrained This is done in the component variable declaration statement The speed and accuracy of the optimization will be increased if the range of the optimization variable is small and its default value is close to optimum value Care must be taken to ensure that the optimum value lies inside the constraint or domain of the optimization variable Simply making the domain large enough will take care of this requirement For instance specifying a radius that lies within the range of 0 to 100 is a safe range within which the optimum radius will lie CircularCylinder isc program main Create the Cylinder component and its variables CreateComponent Cylinder the circular_cylinder CreateVariable Cylinder radius double 0 0 100 0 01 5 0 CreateVariable Cylinder height double 0 0 20000000 0 0 10 0 CreateVariable Cylinder surface_area double 0 0 10000 0 0 0 0 CreateVariable Cylinder volume double 0 0 10000 0 0 0 0 assign the input value Cylinder volume 800 Declare the objective variable AddoObjective Cylinder surface_area 0 Declare the optimization decision variable AddVarObjective Cylinder radius Start optimization calculation Cylinde
116. scripteditor exe iSCRIPT editor an iSCRIPT code developing environment sysdes exe sequential iSCRIPT executable file No diagnostic information is printed out by this executable a quiet mode compared with Parser exe parser exe iSCRIPT analyzer a sequential isCRIPT executable file which is intended to print diagnostics to an output file during execution iscript_mp exe a parallel iSCRIPT executable file which can calculate the problem in parallel machines SampleScripts contains the iSCRIPT sample problems ata contains the iSCRIPT codes for the ata project doc contains the integrated help files for iSCRIPT Editor unist contains the iSCRIPT uninstall information 2 4 iSCRIPT Files There are two types of iSCRIPT files project files and script files with the file extensions ipr and isc respectively Sample iSCRIPT codes are presented in Chapter 3 In this manual iSCRIPT files and the source scripts developed to solve a specific problem are referred to as aniSCRIPT solution Note iSCRIPT is not case sensitive 2 4 1 Script Files isc An iSCRIPT file contains source code Each file may contain a single program subprogram function or component subprogram or simply lines of codes not wrapped in any particular structure program subprogram or function A component subprogram may be considered as a special type of subprogram that computes or solves equations necessary for the perfor
117. sessing CFD Modeling of Entropy Generation for the Air Frame Subsystem in an Integrated Aircraft Design Synthesis Procedure AIAA Aerospace Sciences Meeting and Exhibit Reno NV Jan 2006 Molyneaux A A Practical Evolutionary Method for the Multi Objective Optimization of Complex Energy Systems including Drivetrains M Sc Thesis Ecole Polytechnique Federale De Lausanne EPFL 2002 Leyland G B Multi Objective Optimization Applied to Industrial Energy Problems Ph D Thesis Ecole Polytechnique Federale De Lausanne EPFL 2002 Stoecker W F Design of Thermal Systems Third Edition McGraw Hill Book Company 1989 136
118. so allows the use of INTEGER 2 or INTEGER 2 keywords Long variables Long variables have values in the range 2 147 483 648 to 2 147 483 647 They are also type INTEGER in FORTRAN The syntax for the declaration also allows the use of INTEGER INTEGER 4 or INTEGER 4 keywords Real variables Holds signed IEEE 32 bit 4 byte single precision floating point numbers ranging in value from 3 4028235E 38 through 1 401298E 45 for negative values and from 1 401298E 45 through 3 4028235E 38 for positive values The syntax for the declaration also allows the use of SINGLE REAL 4 or REAL 4 keywords Double variables Holds signed IEEE 64 bit 8 byte double precision floating point numbers ranging in value from 1 79769313486231570E 308 through 4 94065645841246544E 324 for negative values and from 4 94065645841246544E 324 through 1 797693 1348623 1570E 308 for positive values The syntax for the declaration also allows the use of REAL 8 or REAL 8 keywords 74 A 1 2 Variable Names Variable names may be up to 24 characters long and can be alphanumeric However variables must not begin with a _ or a numeral and must not be separated by spaces Examples of valid variable names include e ii Reynolds No Ma jNo iwhat Total_Value A 1 3 Numbers Conventional decimal notation is used with an optional decimal point and leading plus or minus sign for numbers Scientific notation uses the letter e to specify a power of ten s
119. stem consists of only ome component CreateComponent Rastrigin Models entire system CreateVariable Rastrigin y double 0 0 1 0E14 0 0 0 0 CreateVariable Rastrigin xl double 0 0 10 0 10 0 0 5 CreateVariable Rastrigin xZ double 0 0 10 0 10 0 0 5 AddObjective Rastrigin y 0 AddVar Objective Rastrigin xl AddVar Objective Rastrigin xZ 3 Optimize the system Rastrigin Execute Global maxPopulation 70 Global maxInitialEvaluations 200 Global maxGenerations 10 Global optconvergencelimit 1 0E 7 Global wtationfreq 0 2 Global sampsizepervariable 90 Rastrigin Optimize end program subroutine Rastrigin This subroutine is the model for the component Rastrigin xl xZ pi As Double xl Rastrigin xl xz Rast rigin xZ pi 4 atan 1 0 fpi 3 1415926536 Rastrigin y 20 0 xl xl xZ xZ 1O0 cos Z2 pi xl cos 2 pi xzZ end subroutine Figure 6 2 isCRIPT program for finding the optimum of the Rastrigin function The program is created in a single file rastrigin isc with the following parts Main Program The main program starts on Line 3 with the statement program main and ends on line 31 64 The first part of the main program Lines 9 through 12 indicates a component with three variables x7 x2 and y This problem is considered to be a system consisted of one subsystem with a single component This part of the program also identifies y as the obj
120. t variables Note that the equation to solve an objective function is contained within the component model in iSCRIPT and the result of evaluating the objective function is the objective variable f say as H x 0 based on component variables x x1 X2 Xn of n variables f f x Optimize w r t f Then the component variables are x x1 X2 Xn f or n 1 variables and the component model additionally includes the equation f fx 0 Note that iSCRIPT optimization functions are multi objective and f can be a vector of objectives fi The list of variables in x that are free for optimization Note that the fewer variables have a degree of freedom with respect to optimization the faster the optimization will complete The list of components encompassed within the component which is technically a subsystem to be optimized to perform a subsystem level detailed optimization The optimization command is invoked for the component Note that the model equation for the component may consist of execution and optimization commands for other components resulting in nested optimization loops Care must be taken when setting up nested optimization loops as the time required increases geometrically with the number of nesting The procedure for providing the above information is described in subsequent sections B 2 1 Indicating an Objective Variable An objective variable may be defined using the following syntax
121. t you exit all Windows programs before running this Setup program Click Cancel to quit Setup and then close any programs you have running Click Next to continue with the Setup program WARNING This program is protected by copyright law and international treaties Unauthorized reproduction or distribution of this program or any portion of it may result in severe civil and criminal penalties and will be prosecuted to the maximum extent possible under law Next gt Cancel e Click Next Software License Agreement Please read the following License Agreement Press the PAGE DOWN key to see the rest of the agreement COPYRIGHT The following is a notice of limited availability of the code and disclaimer which must be included in the prologue of the code and in all source listings of the cod Copyright Notice 1993 University of Chicago 1993 Mississippi State University Permission is hereby granted to use reproduce prepare derivative works and to redistribute to others This software was authored by Argonne National Laboratory Group Do you accept all the terms of the preceding License Agreement If you choose No Setup will close To install MPICH you must accept this agreement lt Back Yes No e Select where to install MPICH then click Next Choose Destination Location Setup will install MPICH in the following folder To install to this folder click Next To install to a
122. ted if you do not wish to adopt a program structure for your solution However you must adopt a program structure if you must use subroutines or functions in a solution The details of these procedures are illustrated in sample problems in Chapters 4 and 5 3 2 Using iSCRIPT for Performance Analysis of Engineering Systems iSCRIPT uses a decomposition technique to simulate an engineering system Outside of iSCRIPT the system is first decomposed into several components either physically conceptually or along disciplinary lines The variables of each component as well as the constraints engineering limits material constraints etc to the variables are then identified An example could be the fluid inlet temperature into the Burner component of an aircraft engine This variable may be constrained to temperatures less than 2800K for safe operation of the burner material Operating variables are also identified and are considered as system variables as they do not belong to a particular component An example could be the flight Mach number and altitude at which an aircraft engine is operating In iSCRIPT each component is modeled in a separate subprogram a piece of iSCRIPT code containing the equations of the component The component routine may also access or call other functions subprograms or other component routines as part of its model Interactions between a component and other components may also be included in a component subprogram
123. tem System Optimize command is used instead of System Execute Figure 3 3 summarizes the basic procedures required to optimize an engineering system in iSCRIPT Note that the objective variable the variable that we want to optimize is declared by using the statement AddObjective The optimization variables the variables that are free to be changed during the optimization are declared by using statement AddVarObjective The basic steps are listed below Step 1 Open iSCRIPT Editor by selecting iSCRIPT Editor from the Windows Start menu Step 2 Create a new file in iSCRIPT Editor and save it as an isc file Step 3 Type the words program programname and end program to create a main program structure In the body of the main program a declare all the system s components and their variables To declare a component use the CreateComponent statement To declare a component variable use the CreateVariable statement b declare the overall system and its variable using CreateComponent and CreateVariable statements c assign the input values to the system or component variables by typing ComponentName VariableName expression d declare the objective variable by using AddObjective statements e declare the optimization variables by using AddVarObjective statements f indicate the relations between the system components and the system by using AddSubComponent sta
124. tements g type SystemName Optimize to optimize the overall system Step 4 After the main program type keywords subroutine ComponentName and end subroutine to create the component subroutines for each component In the body of the subroutine a type all the equations modeling the component Step 5 After the component subroutines type the word subroutine SystemName and end subroutine to create the system subroutine In the body of the subroutine 20 a evaluate each of the system s components by typing ComponentName Execute b if there are interactions between the components type the interaction equations between the components c type the overall system performance evaluation equations Step 6 Save the file Step 7 Select Tools gt Run Current Script Project File from the menu to run the iSCRIPT code Step 8 View the output on screen or in the file outputscript txt The details of these procedures are illustrated in sample problems in Chapter 5 21 Create the script file in iSCRIPT Editor Decompose the engineering system Create the main program Create components by using CreateComponent statement Create component variables by using CreateVariable statement Create system by using CreateComponent statement Create system variables by using CreateVariable statement Assign the input value for system or component variables Declare the objective variable
125. tf may be used instead of the write keyword e argi arg2 argN are strings variables or arrays Strings must be enclosed in single quotes e size is a value returned representing the number of bytes or characters written A 8 5 Rewinding a File An open file may be returned to the start of file using the close command The syntax is shown below call rewind unit or ivar rewind unit The following rules govern the process of calling the close command 90 e unit is an integer between 10 and 100 representing the handle for the open file e The output of the rewind command is 0 if successful and 1 if an error occurred The exact error based on the operating system is written in the iSCRIPT log file e frewind may be used instead of the open keyword A 8 6 End of File Function An end of file eof command may be called to determine if end of file occurred during the last call to the read command on a specified file open handle The syntax is as follows ivar eof unit The following rules govern the process of calling the close command e unit is an integer between 10 and 100 representing the handle for the open file e The output of the eof command is 1 if end of file occurred or 0 otherwise e feof may be used instead of the eof keyword A 9 Object Oriented Features and Component Modeling iSCRIPT has object oriented features that allows you to define component objects or structures and variables attached t
126. the iSCRIPT installation folder 4 Open the file HeatRejection ipr 5 Select Tools gt Run Current Script Project File from the menu The optimization calculation requires a few seconds to run The final result is contained in the file outputscript txt The results are compared with those from Stoecker 5 in Table 6 1 below A w Csys Stoecker 202 6 167 9 53812 6 iSCRIPT 202 52 167 95 53829 6 Table 6 1 Comparison of iSCRIPT optimization results for Problem 6 2 with those from Stoecker 5 Details of the optimization process are contained in the file optimize txt The details include the initial values of the optimization variables and the objective function the various realizations of the system the array of viable systems or realizations population of individuals in genetic algorithm parlance by generation or as the optimization progresses and the final results A sample optimize txt file is illustrated in Figure 6 5 below 60 DB optimize_final txt Notepad File Edit Format View Help OBJECTIVES INITIAL VALUES waste_sys cost 220016600 cote OBJECTIVE VARIABLES Optimization Parameters maxInitialEvaluations 2000 maxPopulation 200 maxGenerations 30 optConvergenceLimit 1 000000000000000E 010 mutationfreq 1 000000000000000E 002 nsampsize 70 Evaluations and range 256 Population limit initial trials TRYING variables O O 1000E 01 O0 1000 04 OPTIMIZING INITIAL
127. the pump the cost is C pump 9 0005w 5 3 For the condenser the cost is Condenser 270t 3 5 4 The unknown variable is calculated by the energy balance between the condenser and cooling tower O ies O niin 5 5 O oger 3 7w tA 5 6 This problem assumes that the temperature of the water leaving the cooling tower is equal to the temperature of the water entering the condenser i e there is no temperature change in the pump 45 tower l condlenser 5 7 The iSCRIPT code modeling the above system is developed to be consistent with the decomposition procedure adopted for the design The variables used for the system and each of the subsystems are summarized in Table 5 1 below System Variable Remarks Heat Rejection Coys The total cost of the heat rejection system System w Flow rate of cooling water kg s Components Variable Tower Cower First cost of cooling tower A Area of cooling tower m3 Opwer Heat rejection rate of the cooling tower W big Temperature of water leaving the cooling tower C Pump Gimp Life time pumping cost Condenser Condenser Life time penalty in power production due to elevation of temperature of cooling water O scndenser Heat absorption rate of the condenser W tin Temperature of water entering the condenser C Table 5 1 Components and variables of the heat rejection system The development of the iSCRIPT code follows the gener
128. the same name as the component This subroutine includes statements implementing the equations to compute the surface area and volume Note the way in which component variables are used compared to the procedure for Problem 4 3 in Section 4 4 The script file can be found in the subfolder SampleScripts CircularCylinder Example5 1 of the iSCRIPT installation folder If the program is run the same result as in Problem 4 3 will be obtained 5 4 Writing a Program in Several Script Files iSCRIPT allows you to write a program in several files Actually this is highly recommended when the program contains several components This will make the program more portable and easy to manage in a shared project environment In addition a component file can easily be re used or shared by other systems that have the same component It is also easy to add new components to the current system or 4 modify the available components since the program structure matches the engineering decomposition physical conceptual or disciplinary of the system To link all the script files to a project the user must write a single project file ipr which records the name and path information of all the script files The project file must start with a keyword Project in the first line and each line can only contain the one script file name while the path of each script file must be included with the file names if all of the files are not in the same folder
129. the variable number or expression on the right is true or non zero and the variable number or expression on the right is also true LOGICAL OR Returns a value of 1 or true if either the variable number or expression on the right is true or non zero or the variable number or expression on the right is true NOT Returns a value of 1 or true if the variable number or expression on the right is not true or is zero A 1 8 Expressions and Equations Expressions may be generated as a combination of variables numbers and operators Examples include Re rho U L mu Speed_of_sound gamma P rho 0 5 iparameters_provided Reynolds amp Ma iSCRIPT is case insensitive and free form This also means that empty lines spaces and comments can be included as desired without consequence to the performance of the scripts This also means that programs can be indented and commented appropriately for easy code maintenance A 1 9 Comments Comments may be included in a script using the or symbols Comments may occupy a whole line or be included after an expression In either case all input following a comment symbol is ignored Examples include The parameters of the flow are computed below Re rho U L mu Reynolds no Eqn 1 3 Speed_of_sound gamma P rho 0 5 Sound speed Eqn 1 4 T11 A 2 Arrays Scripts may include arrays of any size or dimensions Arrays may be declared similar t
130. ther Software 8 1 Purpose iSCRIPT provides several methods for interfacing with other third party software This is extremely important in providing functionality as a system of systems tool for engineers The methods in iSCRIPT for interfacing with other software include 1 Open a process and execute the third party executable directly This method is universal and works for any third party software as long as the software can be opened from a shell or the command line 2 Provide direct interface to specific software include Microsoft Excel TTC Technologies INSTED software programs and Database and TTC Technologies AEROFLO multi disciplinary CFD program gt 3 Providing the syntax support necessary to run scripts developed in other environments such as MATLAB directly in iSCRIPT with minimum modification The procedures for executing a third party executable are described in the next section 8 2 Running a Third Party Software or Executable An external or third party executable may be activated in iSCRIPT as if from a command line using the execute command The syntax is shown below call execute executable_filename commandline_argument or iresult execute executable_filename commandline_argument The above rules govern the process of calling the open command e executable_filename is a string representing the executable file The string may include the path if the file is not in
131. thin which the above statement is placed is referred to as the calling program or subprogram e The actual arguments to the function must be declared in addition to any other variable declared within the calling program or subprogram e Atruntime the values of the actual arguments if modified within the function is returned to the calling program or subprogram e Atruntime the value of the functionname as a variable is returned to the calling program or subprogram and used to evaluate the expression to the right of the assignment symbol A 5 6 Return When used in a function or subroutine the return statement acts in exactly the same way as the end function or end subroutine statements A sample of syntax is illustrated below subroutine subroutinename arg1 arg2 argN variable declaration statements return end subroutine 84 In the above syntax on encountering the return syntax the subroutine is ended and rest of the subroutine is not executed Used within decision statements the return keyword may be used to conditionally end the execution of a subprogram when certain outcome has been attained A 5 7 Argument Passing Convention Arguments are passed by reference in iSCRIPT similar to FORTRAN However if MATLAB syntax is selected arguments are passed by value Expressions and literal values are passed by value Global and component variables are also passed by value since they are global variables and do not need
132. tic and algebraic problems in Section 3 1 The steps involved in solving this problem are listed below Problem 4 1 Calculate 2 3 in iSCRIPT 4 2 1 Using a Program Structure 1 Open iSCRIPT Editor by selecting the program from the Windows Start menu 2 Create a new file by selecting File gt New from the menu and save it as an isc file e g math isc 24 In the first line of the file type the word program main In line 2 declare a real number variable x by typing x as real In line 3 type the equation x 2 3 In line 4 type the word end program Run the program by selecting Tools gt Run Current Script Project File from the menu 8 After the program is run the results can be viewed on the screen or in the file outputscript txt located in the same folder in which you saved the math isc file Ao eee p The complete script is presented in Figure 4 2 You can also find the script file math isc in the SampleScripts Math Example4 1 subfolder of the iSCRIPT installation folder iSCRIPT Editor C iScript SampleScriptsWath Example 4 1 math isc DER C Fie Edit Search view Document Project Tools Macros Window Help X DLE HF SR 28 m gh BBA w YVE gt A mathisc M Ch1 ASCII DOS READ REC COL OVR Figure 4 2 iSCRIPT code for the computation of 2 3 using a program structure Let s take a quick took at the
133. tine subroutine Pump Pump cost 2 2E8 Pump w Z end subroutine subroutine Exhaust_sys Pipe Execute Pump Execute Exhaust _sys totalcost Pipe cost Pump cost end subroutine Figure 3 3 General procedures for optimizing an engineering system in iSCRIPT 22 4 1 4 Using iSCRIPT in Simple Programs In this chapter we will introduce the most basic iSCRIPT statements and use them to write some simple programs The simplicity of the sample problems does not limit the usefulness of the basic iSCRIPT procedures Getting Started Using 1SCRIPT for the first time is easy In this and the following sections we will introduce you to the iSCRIPT Editor environment and show you how to develop and run a simple iSCRIPT code To start iSCRIPT Editor use the Start menu to locate the program The default iSCRIPT Editor screen which opens each time you start the program is shown in Figure 4 1 iSCRIPT Editor File Search View Project Tools Macros Help Ce he HEAR eel ecm AaB y Ele n gt 7 Al 1 ASCII DOS READ REC COL OVR 4 Figure 4 1 iSCRIPT Editor opening window The iSCRIPT Editor opening window is similar to most text editing or program development environments Detail editor commands may be accessed by clicking Help on the menu However only the basic command menus required for creating and editing iSCRIPT codes are presented in this manual These menu items are presented bel
134. tions for computing rating the aircraft engine an iSCRIPT component model file This model file includes commands to execute the MATLAB model An input file is created for the MATLAB program and the output from MATLAB is read The output is further used to compute certain quantities such as the total exergy destruction in the engine 71 Procedures for Running this Solution 1 Sinaia Run the MATLAB program in MATLAB to check the model of the aircraft engine Compile the MATLAB program into an executable This step includes issuing the command mcc m PS_conv An executable named PS_conv exe is generated This executable was renamed to PS exe Run iSCRIPT Enter the project file ata ipr at the SCRIPT prompt View the results Note that the model may be optimized further on the high level using the optimization procedures present in the iSCRIPT program as described in Chapter 6 12 9 Conclusions Future development of TTC s scripting language will continue to extend the current capabilities adding more intrinsic functions e g for non linear analysis dynamic analysis solvers etc The consistent motive would be to create an easy powerful modeling and design optimization tool compatible with the scripting languages that engineers use most often A GUI procedure to utilize the power of the scripting language and optimization procedure in graphic block building approach will also be separately available Sample scr
135. type of variables For instance the system can be broken down into operational variables that vary in time and those that do not vary in time Time decomposition decomposes a dynamic problem into a series of quasi stationary problems or a series of stationary time segments 1 1 2 Modeling Modeling of the various subsystems in a large system typically involves software products from different vendors A great challenge in this step is the integration of the different software products Several levels can be identified in the multi level modeling and optimization process Low Level Function Interpreters and Symbolic Language Programs These are tools that allow an engineer to specify the equations and models comprising a component in a mathematical form aggregate these low level models into higher level models through additional mathematical expressions and functions In principle a complete system can be built using low level functions However the procedure is difficult and prone to error Sample tools in this category include spreadsheets such as Excel and programming and scripting languages such as MATLAB Mathematica or Maple Aggregated Component Tools These are pre packaged tools for specific models Examples include engine simulator for computing the thrust and weight of the propulsion system e g Weight Analysis of Turbine Engines WATE or ADVISOR a public domain drive train analysis tool or a heat exchanger program for t
136. using the command are acceptable CreateComponent Burner Engine Burner CreateComponent Burner 66 99 Note that the comma inside the square bracket is omitted when the choice is that of using the syntax without the square bracket File names and paths are usually italicized Times Roman For example in the instruction below the italicized portion represents a file path and name Open the file SampleScripts HeatRejection outputscript txt in the iSCRIPT installation directory 2 Installation 2 1 System Requirements ISCRIPT requires the following e Windows 2000 XP 2003 Mac or Linux e Pentium processor with at least 128MB 2 2 Required Software The following software products may be required in order to use iSCRIPT e iSCRIPT installation file e MPICH for execution of iSCRIPT in parallel Note MPICH is optional It is only needed for parallel calculation If sequential calculation is needed you do not need to install MPICH Obtain these software products as shown below iscript_install exe 1 The installation file for ISCRIPT can be downloaded from http www ttctech com SiteFiles Downloads iscript_install exe mpich nt 1 2 5 exe 2 MPICH developed by Argonne National Laboratory is a freely available portable implementation of MPI a standard for message passing protocol for distributed memory applications used in parallel computing MPICH can be downloaded from http www mcs anl gov mpi
137. wn on the screen when the program is completed You can also find the result in outputscript txt in the same folder as the script file a To create this code you can follow the following procedures in Section 4 2 but type in the program instructions as shown in Figure 4 6 30 1 Open iSCRIPT Editor by selecting the program from the Windows Start menu 2 Create a new file by selecting File gt New from the menu and save it as an isc file e g AircraftDrag isc NIH NN BW from the menu 8 After the program is run the results can be viewed on the screen or in the file outputscript txt which is located in the same folder in which you saved the AircraftDrag isc file The results of the calculation are On the first line of the file type the word program main On line 2 declare a real number variable x by typing x as real On line 3 type the instructions as shown in Figure 4 5 On line 4 type the word end program Run the program by selecting Tools gt Run Current Script Project File Cd 2 0019 X 107 V_res m s 0 8 94 17 88 26 82 35 76 44 70 53 64 62 58 71 52 80 46 89 40 Drag_res N 0 800 3200 7200 12800 20000 28800 39200 51200 64800 80000 Analysis of the lines of the program drag rho 3 4 Cd as real 5 6 V as real v_res 11 drag_res 11
138. writing append data to the end of the file 88 A 8 2 Closing a File An open file may be closed using the close command The syntax is shown below call close unit or ivar close unit The following rules govern the process of calling the close command unit is an integer between 10 and 100 representing the handle for the open file The output of the close command is 0 if successful and 1 if an error occurred The exact error based on the operating system is written in the iSCRIPT log file fclose may be used instead of the open keyword A 8 3 Reading from a File or the Keyboard The syntax to read from an open file or the standard input usually the keyboard is shown below call read unit format arg arg2 argN call read unit format arg arg2 argN or A read unit format isize A fscanf unit format isize The following rules govern the process of calling the read command unit is an integer between 10 and 100 representing the handle for the open file No unit specified units 1 2 5 or 6 refers to the keyboard format is a string representing the read format The format string may be omitted simply provide an empty or an used instead The format string when read is currently ignored but is accepted for compatibility with future versions of iSCRIPT The MATLAB format specifiers are accepted fscanf may be used instead

User`s Manual - TTC Technologies

Contents

Download Pdf Manuals

Related Search

Related Contents