Home

Machine Model Parameter Determination Final Report

1. Figure 4 3 Spring 2007 Schedule May07 18 Resources and Schedules Page 51 of 65 Task Name Sep 10 06 Sep 24 06 Oct8 08 Oct22 06 Nov 5 06 Now19 06 Dec3 06 Dec 17 06 Dec 31 06 Jyan14 07 Jen28 07 Feb11 07_ Feb 25 07 Mar 11 07 Mar 25 07 TITISI F J WIMISITITIJSI F WIMISITIT SIF WIMISITITISIJF JWIMISITITIS F WIMISITIT IS F WIM S Task 12 Project Reporting Subtask 12a Project Plan Development EE Subtask 12b Design Report Development E Subtask 12c Project Poster Development EE Subtask 12d Project Final Report Development E Figure 4 4 Deliverables Schedule May07 18 Resources and Schedules Page 52 of 65 5 Closure Materials This section contains an evaluation of the project in terms of success commercialization possibility recommendations for additional work what lessons were learned and risk management It also contains a closing summary and the project participants contact information 5 1 Project Evaluation This section explains the degree of success of the project It shows several main functional requirements and their degree of success The success of each requirement is evaluated as one of the following greatly exceeded exceeded fully met partially met not met or not attempted Any feature not evaluated as fully met in
2. Figure 3 5 Home Window The home window GUI m is shown in Figure 3 5 This is the main interface that collects parameters from the user and displays the graphs of various parameters versus time This window can in general be divided into three separate panels The first is the toolbar that appears across the top portion of the display This panel provides many features that are intended to make the software more user friendly These are discussed in Table 3 3 in Section 3 2 3 1 1 The code necessary to implement these features is discussed in Section 3 2 3 1 2 The panel on the left provides space for the simulation parameters while the panel on the right provides space for a graph of each output variable versus time The amount of space allotted to each sub panel varies depending on the number of parameters displayed For every 16 parameters an additional column of parameters is added and an additional 6 8 percent of the display is allocated to parameters For each parameter a sub panel consisting of an edit box and slider bar is provided The edit box consumes approximately 70 percent of the sub panel while the slider bar consumes the remaining 30 percent Under normal display resolutions the slider bar is small enough that only the up and down increase decrease arrows are visible For each slider bar there are default minimum maximum and current values which are found in the parameter_setting data structure The initial
3. yor a gt p dsp uo QejBue OM let 6z lt la xa e yoowd 4 Git 3492 i zz i peeds gt ENEZAI A lt a EAS LAL pe L Ibdda bdd 4 bdda O vio e lt a 5 paed Ipddal eipue amp a 6ue O eui bddx bx E dh S e x O 2 a x E E o 2 iz o bjd bdd Llbddal gt e eae i piper Ip gt gt of px gt A Pi P pipe Zv IX PAX pddx pdx y EE O IX pdx s unpes as IX Pax a dais gt pddx pdx H md a E an gt Qe en opdd j 1 amp lt b amp pe eo e 1X pdx IX pddx pi3 20 SL ZO gius Page 64 of 65 Sample Simulink Block Diagram May07 18 Appendix D Opening Window Detailed Flowchart initialize GUI Prompt user to select machine model from Main Window mport test data m Invalid file format Home Window r Load block diagram C y Load parameters value to GUI User changes parameters Batch Simulation Window Load machine setting fie p batch parameter Plot test data in GU User clicks batch simulation button r Update parameters value in block diagram M r m Batch
4. Wait for user s action User clicks simulate button Executes simulation script Batch Simulation Plot simu nk result in GUI User exit GU eo May07 18 Detailed Flowchart Page 65 of 65
5. 10 100 T pats 10710 1005 4 4 10 10 10 10 1000 1000 60 401 no default value VE GI on T Elt GEL Atenlt 2 Rt Xt Rl1 XI MVA machine MVA system fbase LS xis ely Sy Dy Zeby Bede ews 100 Tg 087 4303554 34 04 Lied 295 4 AZ Ay alby 205 5 22 529 Q0 0 previous_value 0 current_value Ty 03y 2095 3 Op Lil 497 AZ e O 305 35 Uy Oy 100 32 09 1 05 100 32 09 1 05 100 32 09 1 05 Mi LL LOs Oy sly 0 27 120 100 60 30 004 May07 18 Sample GUI Setting File Page 61 of 65 100 32 09 1 05 100 32 09 1 05 100 32 09 1 05 Al 1 ie Oe ex ieee 0 gle Or d 120 08 Sinus 30 004 Lara display name Tpdo Tppdo Topgqo H re Eat Dus PPM Lpd Lppd bl S1 S12 Ra Rcomp Xocomp Po n Vstp Ogen Vstp Vvi Vstp Pgen Rjot Ogen RJjot Vl Rjet Pqgen F1ET Ogen Flt Vl FIt z Tup Vstep V Vstep VER RICE as UT flt on TIT Elt off Z2fault EE xe BA X1 MVA mach MVA sys fboase Tstop Step Size fress description L SoLo yellow yellow yellow yellow yellow yellow yellow yellow yellow yellow yellow yellow yellow yellow yellow white w
6. 5 era et hend Rent ose UE S d tee Race aed ee tha 56 5 5 1 Anticipated potential risks and planned management 57 5 5 2 Anticipated risks encountered and success in management 57 5 5 3 Unanticipated risks encountered attempts to manage and success 57 May07 18 Page ii of vi 5 5 4 Resultant Changes in Risk Management Made Because of Encountered Uranticrpated RISKS a Qao sas cod tu d moe db vices bere Det nasa s ue REN 57 5 6 Project Team Information ia 57 5 7 Client SfOFIatlONi x AA E a is reote Sa ae 58 5 8 Advisor Informat A oou ttai cde lote mE e eet 58 5 0 Closing SUMMA Vii 58 Appendix A Sample Test Data oe peer A cde Gall a de tes 60 Appendix B Sample GUI Setting Filete ii ii 61 Appendix C Sample Simulink Block Diagram eee 64 Appendix D Detatled Plow CW ar tisse sites ecase thse tore ira 65 May07 18 Page iii of vi List of Tables Table 1 1 Initial Assumptions and Justifications oonoccnocccnocnnaoncnoncnoncnnnn nono conocio no ncnnnncnnnnos 3 Table 1 2 Initial Limitations and Justification sese 4 Table 2 1 Product Requirements and Features ede dias colas bei Poetas eee depu 5 Table 3 1 Sample Test Data Formal aio 30 Table 3 2 Data Structures and their Fields deste eio docto bo e aide da Peta efus 33 Table 3 3 Product Requirements and Features eese 38 Table 4 1 Origi
7. to Simulating The batch simulation window See Figure 3 8 appears and the program suspends execution until the user presses either the OK or cancel button After this the program resumes execution 1f the cancel button was pressed the program returns to the main function and does not simulate If the OK button was pressed a wait bar is displayed and the function collects the parameter minimum value maximum value and number of simulations from the workspace After this an array of parameter values for the parameter that will be varied is created and all of the parameters are placed in the workspace After the simulation has completed the data is plotted The program repeats this for the desired number of simulations May07 18 Detailed Design Page 40 of 65 EDTITUTENESSO Lini Settings Simulations ka Maximum Value 5 Minimum Value 2 OK Cancel Figure 3 8 Batch Simulate Window 3 2 3 1 2 5 Open Button Callback Function The open button callback function starts by opening the standard windows Open File window This window is shown in Figure 3 9 The user has the option to either import test data from an Excel spreadsheet or load a saved case in the form of a Matlab mat file May07 18 Detailed Design Page 41 of 65 Figure 3 9 Open Window After the user has selected a file the program attempts to open it as an Excel file if it succeeds the data is plotted if it fails the p
8. May07 18 Detailed Design Page 28 of 65 ny Select machine model M y o _ Fd A Select machine test x y ra ON N Batch simulation Edit parameters x M vi Continue Simulate and plot result NI F Quit wy Figure 3 2 Simplified Flow Chart 3 1 1 2 Graphical User Interface The graphical user interface is intended to provide the user with an easy means of entering parameters and viewing selected simulation results It also provides several features such as the ability to undo or redo certain actions and import or export data that make it much more user friendly It can be broken into three basic components the opening window main_window m the home window GUI m and the batch simulation window batch_simulation m The basic purpose of the graphical user interface is to collect parameters from the user store them in the Matlab workspace execute the simulation script collect the arrays of simulation data and plot them 3 1 1 3 Simulink Models The Simulink models are block diagram representations of the state space equations necessary for time domain simulations It is very important that all constants be entered in as variables to ensure that they can be easily changed and that they will be changeable from the graphical user interface Selected parameters are stored as arrays in the Matlab workspace The standard practice has been to use embedded Matlab fun
9. 2 of o Task 2d Technology Sec o 7 2 7 Task Sa Montficaton of Design Requirenens 1 2 AA Taro Design Process O BOI o Task 3e Documentation ofDexign 3 9 8 _2 Task 4 Prototype implementation __ v a 1s 89 Task S ProtoiypeTesting s 5 2 9 s Task SP Tes Develop 9 o o Task 3e Test Brecuton LT A CEE 9 AA ask Se DocumenarioorTesig 9 9 5 Task 60 Feedback inpleneniation 7 Task 7 Ena Product Design jua 1 EN hu N N No Task 8b Implement End Product User Interface o y A a mln a wl ve u Ln 4A S Prada Tes Planning 5 Task 9b Tes Development 2 Task 9d Test Evaluation 2 Tast 9e Documentation TEE AA Task 10 End Product Documentation Development Task 8 End Product Implementation Task 8a Implement End Product Block Diagram Task 9 End Product Testing May07 18 Resources and Schedules Page 48 of 65 Task Mark Adam Jared Yu Task Task 11 End Product Demonstration a z w e al Task Ta Denonsaton Pang H3 i 58 Task 11 Faculy Advisors Demonstration __ 2L E Task Tie len Demonstration 1 LE E Task d Indusval Review panel Demonstration 1 1 E Task 12 Project Repo
10. Display on Graphs domain data stored in Microsoft Excel spreadsheets in a format specified by the project team and plot them on the same graphs as the simulation data to previous simulation of each slider bar of each slider bar from a previous session on all graphs May07 18 Detailed Design Page 38 of 65 Table 3 3 Continued Feaure supilemenal nendecFuncionaley Feature Supplemental Intended Functionality Handle Simulation of Required The graphical user interface shall be able to Different Tests handle models capable of simulating different events Handle Multiple Models Required The graphical user interface shall be capable of handling a library of any number of models The project team shall enter at least 5 models into this library Supplemental The user should be able to pan graphs Save Graphs as Matlab Supplemental The user should be able to save graphs as Figures Matlab figures Export Simulation Data to Supplemental Excel Spreadsheet The user should be able to save simulation data as an Excel spreadsheet in a format that is the same as the imported data Batch Simulation Supplemental The user should be able to choose a single parameter specify a minimum and maximum value and any number of simulations and have the program iterate that many times between those two points and graph the simulation data Hold Previous Zoom Level Supplemental When test data has been imported the Graph s zoo
11. IZ The Matlab Workspace sxc sated idas 30 3 1 3 Settna Files e 30 3 1 4 POSE Date m 30 3 1 5 Interaction Between Components and Flow of Information 31 3 2 Graphical User Interfaces cia 31 3 2 1 Data Structures aussi ae 32 2 22 Opening WIDOONM ta tl diia 34 3 2 5 Home W INOW unie ee ie Mus o eed dodi eura 36 3 2 4 Batch Simulation WI Wii A eee 44 3 3 Si n tities edi aaa Puce eod cu ea aa uds a a e EES 44 3 4 Simulink Models iii 44 Resources and Schedules a A dt Ad eee 46 4 1 IRESOUIDE Requirements dan 46 4 1 1 Personnel Effort Requirements tectis teinte tris a e pe nea tete eH ee eee dh 46 4 1 2 Other Resource Requirements sido lidia 49 4 1 3 Financial Requirements Ra 50 4 2 Veris In R a E R R AN 50 Closure Materials e dd eb s escis 53 5 1 Project EvaltatiO m na a 53 5 2 sCommercialization Possibility sensi enana e aa ees 55 5 3 Recommendations for Additional Work eee 55 5 4 Lessons Learned aite etr dU e tet edes dese 56 5 4 1 What Went Well ioco aident ai ia Ban ptas 56 5 4 2 What Did Not Go Well ient rtt speso trita rad ee e eso eed eee Qin 56 5 4 3 What Technical Knowledge Was Gained sess 56 5 4 4 What Non Technical Knowledge Was Gained sss 56 5 4 5 What The Team Would Do Differently If The Team Had To Do It Over Again 56 3 9 Risk and Risk Management
12. Open previously saved parameter data from a different machine Purpose To verify that the program works properly after the user attempts to import parameter data from a machine with the incorrect number of parameters Procedure Use the open button to import previously saved parameter data Check to see if the exception is handled the user warned and that the program continues to run properly Expected Result The exception will be handled the user will be warned and the program will continue to run properly Actual Result R2006a Same as R2006b R2006b Non detailer message pops up Action Required Reword error message to say Error loading parameter settings or parameter settings for wrong machine Completed Yes 2 1 5 2 2 5 Edit Parameters Values Using Edit Boxes and Slider Bars Test Case Check default values for all edit boxes Purpose To verify default values match data in setting file Procedure Open the home window and check that the default values listed for each parameter are the same as those in the setting file Expected Result Default values in edit box match those in the setting file Actual Result R2006a Same as R2006b R2006b Setting files match Action Required none May07 18 Approach and Results Page 18 of 65 Completed Yes Test Case Enter valid value in edit box Purpose To verify that the edit boxes have been properly implemented Procedure Enter various valid numerical value in
13. 13 5 17 2 40 Diagram Task 6b Implement End Product User Interface 5 14 5 16 40 Task 9 End Product Testing 16 18 14 14 62 Task 9a Test Planning 2 2 4 0 Task 9b Test Development 0 2 2 4 Task 9c Test Execution 2 2 0 3 10 Task 9d Test Evaluation 4 2 0 0 6 Task 9e Documentation of Testing 5 10 amp 7 30 Task 10 End Product Documentation 5 6 9 2 22 Task 10a Development of End User 5 6 9 2 22 Documentation Task 11 End Product Demonstration 9 8 5 4 26 Task 11a Demonstration Planning 7 6 4 3 20 Task 11b Faculty Advisor s Demonstration 0 1 0 1 2 Task 11c Client Demonstration 1 1 0 0 2 Task 11d Industrial Review panel 1 0 1 0 2 Demonstration Task 12 Project Reporting 25 33 19 16 93 Task 12a Project Plan Development amp 10 6 9 33 Task I2b Design Report Development 4 6 9 1 20 Task 12c Project Poster Development amp 8 0 20 Task 12d Project Final Report Development J 9 4 2 20 Project Total 160 186 164 157 667 May07 18 Resources and Schedules Page 47 of 65 Table 4 2 Actual Labor Usage Total Task la Problem Definition Completion of o op op o Task Ib End Users and End Uses Identification o J of he 2 Task 1c Constraint Identification of 2 of X 3 2 3 2 Task 2 Technology and Implementation Considerations and Selection Task 2a Identification of Possible Technologies 9 9 IA Task 20 Identification of Selecion Cera I 9 IA Task 2e Technology Research 2
14. Case Batch simulation with the number of simulations set to a non integer number Purpose To verify that batch simulation has been properly implemented Procedure Click batch simulation button in lower left corner of home window choose a single parameter enter 3 25 into the simulations box and click OK Expected Result Three simulations are run and the graphs appear properly with the correct legend Expected Result Three simulations are run and the graphs appear properly with the correct legend Actual Result R2006a Same as R2006b R2006b Any number of simulations that are a non integer are rounded to the nearest integer May07 18 Approach and Results Page 24 of 65 Test Case Batch simulation with the number of simulations set to a positive integer number Purpose To verify that batch simulation has been properly implemented Procedure Toggle a breakpoint in the home screen code Click batch simulation button in lower left corner of home window choose a single parameter enter 4 into the simulations box and click OK Check to see that the correct parameter is being varied and that the results are plotted correctly with the correct legend Expected Result Four simulations are run and the graphs appear properly with the correct legend Expected Result Four simulations are run and the graphs appear properly with the correct legend Actual Result R2006a Preforms correctly R2006b Preforms correctly Action R
15. Result R2006a Redo button performs correctly R2006b Redo button performs correctly May07 18 Approach and Results Page 20 of 65 Action Required Identify undo redo disabling code and check for uniformity Completed Yes 2 1 5 2 2 7 Zoom and Pan Plots Test Case Zoom in on graphs without test data Purpose To verify zoom in feature is implemented properly Procedure Run simulation without test data loaded zoom in or out on the graphs by pressing the zoom in button and using the left mouse click Zoom out using the shift left click simulate one more time after zooming in or out to ensure graphs reset properly Expected Result Graphs zoom in and out properly graphs reset properly after each simulation Actual Result R2006a Works correctly R2006b Works correctly Action Required none Completed Yes Test Case Zoom in on graphs with test data Purpose To verify zoom in feature is implemented properly Procedure Run simulation with test data loaded zoom in on the graphs by pressing the zoom in button and using the left mouse click Zoom out using the shift left click simulate one more time after zooming in or out to ensure graphs reset properly Expected Result Graphs zoom in and out properly graphs do not reset after each simulation Actual Result R2006a Works correctly R2006b Works correctly Action Required none Completed Yes Test Case Zoom out on graphs without test data Purpose To v
16. graphs as the simulation data to previous simulation of each slider bar of each slider bar from a previous session The user shall be able to zoom in and out on all graphs May07 18 Approach and Results Page 5 of 65 Table 2 1 Continued Required or Feature Supplemental Intended Functionality Handle Simulation of Required The graphical user interface shall be able to Different Tests handle models capable of simulating different events Handle Multiple Models Required The graphical user interface shall be capable of handling a library of any number of models The project team shall enter at least 5 models into this library Supplemental The user should be able to pan graphs Save Graphs as Matlab Supplemental The user should be able to save graphs as Figures Matlab figures Export Simulation Data to Supplemental Excel Spreadsheet The user should be able to save simulation data as an Excel spreadsheet in a format that is the same as the imported data Batch Simulation Supplemental The user should be able to choose a single parameter specify a minimum and maximum value and any number of simulations and have the program iterate that many times between those two points and graph the simulation data Hold Previous Zoom Level Supplemental When test data has been imported the Graph s zoom level should not change after the simulate button is pressed Graph Previous Simulation Supplemental Data The previous s
17. rexs exciter e Unexcited gensal model e Gensal with exstl exciter e Gesnal with exstl exciter and HVGOV governor 2 1 5 1 1 2 Items Included in Model Testing The following items were included in model testing e Verification of default edit box values e Simulation with all parameters set to default values e Verification of minimum edit box values e Verification of maximum edit box values e Verification that varying parameters produces different simulation results e Comparison of simulation data with PSLF simulation results 2 1 5 1 1 3 Items Excluded from Model Testing The following items have been excluded from model testing e Testing to determine what values of certain parameters give numerical errors e Testing to determine if certain combinations of parameters give numerical errors 2 1 5 1 2 Test Cases This section summarizes the tests that were performed for each model All tests not related to verifying edit box and slider bar values were performed on Matlab versions R2006a and R2006b The test procedure outlined in Section 2 3 6 describe the process that the Iowa State team feels would verify that the models produce accurate numerical results Because the Iowa State team does not have access to the software necessary to complete this kind of testing it will be completed by the team at General Electric The exact procedure used was left to their judgment The test case defines what was tested The reason for testing
18. scripts are very flexible the standard practice has been to use a switch statement that switches based on the test selected The initial conditions corresponding to this test are chosen in this statement In the lines that follow more initial conditions are calculated using algebraic expressions The last line executes the Simulink model It is very important that the variables used in the simulation script exactly match the variables used in the Simulink models and GUI setting files It should be noted that it is not necessary to use Simulink block diagrams The simulation script can be used to build a state space model of the system being studied Any one of the Matlab differential equation solvers can be used to perform the simulation 3 4 Simulink Models The Simulink models used for simulations are also very flexible They can be models of any dynamic system and do not need to be from the power industry The standard practice has been to use block diagrams for the generator windings exciters voltage May07 18 Detailed Design Page 44 of 65 regulators turbines and governors and imbedded Matlab code for the algebraic equations corresponding to the electrical network The simulation time and step size need to be adjustable via the home screen thus the appropriate settings in the configuration parameters dialog box Simulation gt Configuration Parameters need to be changed to variables Variables will be used in place of constants
19. sent by General Electric Expected Result All maximum edit box values for all models should match the data sent by General Electric Actual Result R2006a Same error as R2006b except for some computers there is an error Reference to non existent field undo toolbar button When running a simulation R2006b Each model simulates with parameters set to default However if you open a new model without first clearing the workspace there is a simulation error Action Required Insert command to clear workspace variables on main_window m startup Investigate fatal error on R2006a if time permits however error does not occur on majority of computers After further inspection it it appears that this only occurs on corrupt Matlab installations and is not a problem with the GUI Completed Yes 2 1 5 1 2 3 Verification of Minimum Edit Box Values Test Case Verify minimum edit box values May07 18 Approach and Results Page 10 of 65 Purpose To verify that the default minimum edit box values match the data sent by General Electric Procedure Open each model from the opening window and use the slider bar to set all values to the minimum Compare the edit box values to the data sent by General Electric Expected Result All minimum edit box values for all models should match the data sent by General Electric Actual Result R2006a All included models displayed the correct default value R2006b All included models displa
20. simulate button Compare the simulation results with the PSLF results Expected Result All plots will match within a reasonable tolerance Actual Result To be completed by GE Test Case Simulate all models with the parameters set at a value other than its default and compare to PSLF data May07 18 Approach and Results Page 12 of 65 Purpose To verify that all models produce correct numerical results at a value other than their default value Procedure Use PSLF to prepare test data corresponding to a case in which the all parameters would not be set at their defaults Export this data to a Microsoft Excel Spreadsheet whose format matches the one described in the detailed design Open each model from the opening window import the test data Change each edit box to the correct value Press the simulate button Compare the simulation results with the PSLF results Expected Result All plots will match within a reasonable tolerance Actual Result To be completed by GE 2 1 5 2 Functional Testing The objective of functional testing is to verify that The software met the functional requirements specified in the project description and design report All features required for each functional requirement are implemented correctly All features function properly in the possible operating environments 2 1 5 2 1 Scope and Approach The scope of functional testing is to verify that the functional requirements and sets of featur
21. spent time above and beyond adding additional models and tests The formulation of setting files for the models and tests allows engineers to include additional models in the future with ease The batch simulation feature was not only a supplemental component included near the project completion but it also allows for a small amount of automation for any parameter for any number of simulations This will allow the engineer to use their time more efficiently The project as a whole is evaluated as very successful because each requirement was at least fully met and every supplemental requirement will save the engineer additional time and therefore money for the company Additionally the team has provided an assessment of the success of each milestone This appears in Table 5 2 Overall the team feels that they were very successful and that all of the milestones were very successful The team feels that had the client s needs been more clearly determined at an earlier date the second semester implementation might have been smoother and the research though well executed was not fruitful A great deal of time was spent on implementation and the quality of the documentation suffered as a result Testing took much longer than anticipated and the end product reviews though positive did reveal some improvements that could have been made May07 18 Closure Materials Page 54 of 65 Table 5 2 Milestone Evaluation Milestone Relative Evaluation Result
22. this is described in the purpose Procedure outlines how the test were performed The expected result explains what result was considered completely successful The actual result describes what happened when the test was executed and what modifications were necessary to achieve the expected result If the expected result was not achieved and no attempt to remedy the situation was made a justification is given May07 18 Approach and Results Page 9 of 65 2 1 5 1 2 1 Verification of Default Edit Box Values Test Case Verify default edit box values Purpose To verify that the default edit box values match the data sent by General Electric Procedure Open each model from the opening window and compare the default edit box values to the data sent by General Electric Expected Result All default edit box values for all models should match the data sent by General Electric Actual Result R2006a All included models displayed the correct default value R2006b All included models displayed the correct default value Action Required none Completed yes 2 1 5 1 2 2 Simulation with All Parameters Set to Default Values Test Case Simulate all models with the parameters set at their default values Purpose To verify that all models will simulate with the parameters set at the default values Procedure Open each model from the opening window and use the slider bar to set all values to the maximum Compare the edit box values to the data
23. 0 6 Task Ib End Users and End Uses Identification 0 0 2 2 4 Task Ic Constraint Identification 0 2 0 2 4 Task 2 Technology and Implementation 3 4 2 3 12 Considerations and Selection Task 2a Identification of Possible Technologies 0 2 0 2 4 Task 2b Identification of Selection Criteria 1 0 1 0 2 Task 2c Technology Research 2 2 0 0 4 Task 2d Technology Selection 0 0 1 1 2 Task 3 Prototype Design 22 28 23 31 104 Task 3a Identification of Design Requirements 1 3 0 4 amp Task 3b Design Process 15 15 11 17 58 Task 3c Documentation of Design 6 10 12 10 38 Task 4 Prototype Implementation 23 18 16 23 80 Task 4a Implement Prototype Block Diagram 10 11 9 10 40 Task 4b Implement Prototype User Interface 13 7 7 13 40 Task 5 Prototype Testing 11 18 20 11 60 Task 5a Test Planning 2 4 2 0 amp Task 5b Test Development 0 2 0 4 6 May07 18 Resources and Schedules Page 46 of 65 Task Mark Adam Jared Yu Task Total Task 5c Test Execution 2 4 2 2 10 Task 5d Test Evaluation 0 0 4 2 6 Task 5e Documentation of Testing 7 amp 12 3 30 Task 6 Feedback 0 1 3 8 12 Task 6a Client Feedback 0 0 0 8 Task 6b Feedback Implementation 0 1 3 0 4 Task 7 End Product Design 24 29 29 20 102 Task 7a Identification of Design Requirements 1 0 3 0 4 Task 7b Design Process 15 20 12 13 60 Task 7c Documentation of Design amp 9 14 7 38 Task 8 End Product Implementation 18 19 22 21 80 Task 6a Implement End Product Block
24. 27 Task 1 Project Definition 31 Task 2 Technology and implementation considerations an 386 Task 3 Prototype Design 40 Task 4 Prototype Implementation 43 Task 5 Prototype Testing 49 Task 6 Feedback 53 Task 1 Project Definition 57 Task 2 Technology and implementation considerations an 62 Task 3 Prototype Design 66 Task 4 Prototype Implementation 69 Task 5 Prototype Testing 75 Task 6 Feedback Figure 4 2 Fall 2006 Schedule ID Task Name S F IWIM S TI T S FIWIM S TITIS F WIM S TIT IS F IWIM S T Dec 31 06 Jan 14 07 Jan 28 07 Feb 11 07 Feb 25 07 Mar 11 07 Mar 25 07 Apr 8 07 Apr 22 07 May 6 07 T 1 Task 7 End Product Design 5 Task 8 End Product Implementation 8 Task 9 End Product Testing 14 Task 10 End Product Documentation 16 Task 11 End Product Demonstration 21 22 Task 7 End Product Design 26 Task 8 End Product Implementation 29 Task 9 End Product Testing 35 Task 10 End Product Documentation 37 Task 11 End Product Demonstration 42 43 Task 7 End Product Design 47 Task 8 End Product Implementation 50 Task 9 End Product Testing 56 Task 10 End Product Documentation 58 Task 11 End Product Demonstration
25. 4 1 053 1 912173 0 0 03 1 024 1 343 0 999919 0 0 04 1 024 1 354 1 000052 0 0 05 1 024 1 362 0 999749 0 0 06 1 025 1 370 0 999986 0 0 07 1 023 1 340 1 000011 0 0 08 1 024 1 311 1 000044 0 0 09 1 024 1 336 0 999877 0 0 10 1 024 1 341 0 999898 0 0 11 1 024 1 337 1 000077 0 0 12 1 024 1 321 0 999906 0 0 13 1 024 1 327 0 999855 0 0 14 1 024 1 352 0 999912 0 0 15 1 024 1 351 0 999923 0 0 16 1 024 1 308 0 999853 0 0 17 1 024 1 320 0 999923 0 0 18 1 024 1 324 0 999948 0 0 19 1 024 1 334 0 999877 0 0 20 1 024 1 330 0 999842 0 0 21 1 024 1 315 1 000073 0 0 22 1 024 1 332 1 000003 0 0 23 1 024 1 329 0 999991 0 0 24 1 024 1 358 0 999722 0 May07 18 Sample Test Data Page 60 of 65 clear parameter_setting parameter_setting struct workspace_name Epdo t tepuo tspgo h q t Ttig Ae aas Lpd ippd Lil1 si sl2 rya rfoomp compg scs Pgen Vstep Qgen Vstep V1 Vstep Pgen Ld Reject Qgen Ld Reject V1l Ld Reject Appendix B Sample GUI Setting Poen FIT OGOR Flt yl FIt Tup Vstep Vup Vstep Treject Ld Reject Tstop Step size pda min_value Ay xl Oly ep Oy adp Spee aly sly ely 04d uy 04 0 0 1000 OF tre 0 1000 0 Oy 1000 Orasa O 100 45 D eon 050 05 us Oy Oy Qu 0 Lk d 0 001 max_value Oy 287 wk Gp Ly SOL el 1000 1000 1 057 a 1000 1000 1 05 1000 1000 1 05
26. 6 of 65 3 Detailed Design This section is intended to provide the reader with an in depth background in the program s structure and design Selected components are referenced in the appendices of this document 3 1 Design Overview The software was developed in the Matlab programming language and was designed for maximum flexibility As a result of the modular design the structure can be broken into several independent components An overview of each of these components appears in the following Section 3 1 with a more detailed discussion of each in sections 3 2 through 3 4 The Matlab Graphical User Interface Development Environment GUIDE toolbox was used to generate the base code for the three main GUI components Code generated using GUIDE contains a main function that executes the create function and processes the user inputs A callback function is provided for each object button drop down menu list box edit box slider bar etc The layout function creates the GUI from standard Matlab components A create function is also provided for each object The callback functions allow the programmer to create code that will execute each time that a button is pressed text is selected or information entered The callback functions allow for the programmer to create code that will execute when an object is created The create functions were not used for this project 3 1 1 Program Components The program can be broken down into three ba
27. A PARO US A ii 1 1 1 3 Recommendations for Additional Work eese 1 1 2 AXcknowledger ents uiui ose ates REESE ascents ES IUS A R 1 1 3 as men osos dotis ebd dello deoa a de cope UR CEE 2 1 3 1 General Problem Statement ctas 2 1 3 2 General Solution Approach erstens eite tained en sesta ta sopas esed ee toas 2 1 4 Operating EAVITOnNMENt seon RES ERREUR NI Sua ets E i 2 1 5 Intended Users and USES i i ee M as 3 1 5 1 Intended EISet s Lotes eat EAS eee nee ee 3 1 5 2 Intended Usuario e cte nas hess aain 3 1 6 Assumptions and Limitations eese eene entente eene ettet nette tenis 3 1 6 1 Initial A SS UFTIDBLOTIS ds 3 1 6 2 A A a a ia 3 1 7 Expected End Product and Other Deliverables eee 4 1 7 1 Simulink Block Diagram Piles coria 4 1 7 3 Graphical User InterfaCe niniin a a ias 4 1 7 4 Project Doc m ntati n ne aaa tet eee toties E S 4 2 Approach and Results 5 2 1 Approach d 5 Zeal Functional Requirements ouod ta ottiene ete eae aa 5 2 1 2 Desion C ons traits o ed ie dte ceu 6 2 1 3 Technical Approach Considerations eene 7 2 1 4 Implementation Process Description eese 8 2 1 5 End Product Testilig sii iren te ninia 8 2 2 Project End Result ee A a EA aie BAG 26 SMEBDTCCUME PIRE LE UT 27 3 1 DESIST OVER die 27 May07 18 Page i of vi 3 1 1 Program Component tds 27
28. Machine Model Parameter Determination Final Report May07 18 Client General Electric Faculty Advisor Chen Ching Liu Team Members Jared Kline Adam Wroblaski Mark Reisinger Yu Chan Disclaimer This document was developed as a part of the requirements of an electrical and computer engineering course at Iowa State University Ames Iowa This document does not constitute a professional engineering design or a professional land surveying document Although the information is intended to be accurate the associated students faculty and Iowa State University make no claims promises or guarantees about the accuracy completeness quality or adequacy of the information The user of this document shall ensure that any such use does not violate any laws with regard to professional licensing and certification requirements This use includes any work resulting from this student prepared document that is required to be under the responsible charge of a licensed engineer or surveyor This document is copyrighted by the students who produced this document and the associated faculty advisors No part may be reproduced without the written permission of the senior design course coordinator Submitted May 2 2007 Table of Contents Machine Model Parameter Determination Final Report eene 1 I Jdntroductiot io ede i Gretta ere ae cate ee 1 1 1 Executive Summar niga 1 1 1 1 Project ACtVIE S ota ile ies 1 LA
29. Opening window closes properly Actual Result R2006a Closed properly R2006b Closed properly Action Required none Completed Yes 2 1 5 2 2 2 Select Machine Model to Simulate From Opening Window Test Case Select machine model from opening window Purpose To verify the function of selecting machine model from opening window has been correctly implemented Procedure Select each machine model in list box and click open button in opening window Expected Result Home screen corresponding to the selected model should appear correctly and opening window should close after the home screen appears Actual Result R2006a Opened properly R2006b Opened properly Action Required none Completed Yes 2 1 5 2 2 3 Import and Export Test and Simulation Data Test Case Import test data from a Microsoft Excel file Purpose To verify that the program can import test data from a properly formatted Excel spreadsheet for each machine test Procedure For each machine test click the open button in the upper right corner of home window browse for the test data and click ok Test data should be in Excel format specified in the detailed design section Expected Result Test data is plotted correctly in the output panel of home window for each machine test Actual Result R2006a Opened Properly May07 18 Approach and Results Page 15 of 65 R2006b Opened properly Action Required none Completed Yes Test Case Import i
30. Table 5 1 is discussed below The testing supporting that these features have been properly implemented appears in Section 2 1 5 Table 5 1 Project Task Evaluation Feature Required or Supplemental Evaluation Edit Box Required Fully met Slider Bar Required Fully met Interface with Simulink Required Fully met Graph Simulation Data Required Fully met Import Data from Excel and Required Fully met Display on Graphs Undo Redo Actions Required Fully met Save Current Session Required Fully met Load Previous Session Required Fully met Zoom in and Out on Graphs Required Fully met Pan Graphs Supplemental Fully met Handle Multiple Models Required Exceeded Handle Simulation of Required Exceeded Different Tests Save Graphs as Matlab Supplemental Fully met Figures Export Simulation Data to Supplemental Fully met Excel Spreadsheet Batch Simulation Supplemental Exceeded May07 18 Closure Materials Page 53 of 65 Feature Required or Supplemental Evaluation Hold Previous Zoom Level Supplemental Fully met Graph Previous Simulation Supplemental Fully met Data Color Code for Edit Box Supplemental Fully met Changed Status for Certain Supplemental Fully met Parameters to Active or Inactive Based on Currently Selected Test The tasks handle multiple models and handle simulation of different tests were both fully completed to the requirements of the project plan However the team has
31. Users will be able to quickly adjust all machine model Client parameters using up and down arrows as well as text boxes Time required to obtain accurate model parameters will not Client exceed four hours Completed project will include software user manual test Design plan results and this final report The output graphs contained in the GUI will update by Design buttons contained within the GUI itself 1 7 Expected End Product and Other Deliverables This section lists all items that are supplied to the client as part of the end product 1 7 1 Simulink Block Diagram Files Fully functional Simulink files of all machine types for which the group has block diagrams for are included in the software package These files conform to the limitations described in Table 1 2 1 7 2 Setting Files Setting files for each generator type will be included in the software package These setting files tailor the layout of the graphical user interface for each model type 1 7 3 Graphical User Interface A graphical user interface is provided in the form of a Matlab m file It interacts with the included Simulink models and setting files The source code will also be provided 1 7 4 Project Documentation The project team will deliver electronic and bound copies of the final reports and end user manual to ensure that the client understands the software design and its proper and most effective use as well as the design p
32. ail douglas welsh ge com Fax 518 385 3165 Teleconference 877 653 7365 Pass Code 191535 Secondary contacts Dan Leonard Machine Testing Jaun Sanchez Software Team 5 8 Advisor Information Professor Chen Ching Liu 1117 Coover Hall Iowa State University IA 50014 Phone 515 294 4763 Email liu iastate edu 5 9 Closing Summary General Electric engineers currently spend three to five days comparing simulation models to the test results from only one of their generators This process currently involves the engineer changing a large number of inputs many times in complex text based simulation software The goal of this project was to create a fast and highly usable program that can simplify and accelerate this process As has been extensively May07 18 Closure Materials Page 58 of 65 shown above the team has created just such an interface allowing engineers to import and export data sets from their physical testing perform a variety of simulations with many time reducing enhancements and quickly determine the parameter values for the generator By reducing the time needed to run the simulation vary the parameters and compare the results to test data from several days to half a day the company can save hundreds of thousands of dollars in man hours each year May07 18 Closure Materials Page 59 of 65 Appendix A Sample Test Data Time sec Vt pu Efd pu Spd pu Ifd pu 0 01 0 838 0 080 1 911961 0 0 02 1 02
33. ant Importance Score Score Problem Definition End Product Implementation End Product Testing End Product Demo 1096 90 9 5 2 Commercialization Possibility This program is very flexible and has uses in virtually every case where the parameters needed for accurate computer modeling are difficult to determine The project team purposely did not make any decisions that would limit the program to any one field However because of licensing difficulties associated with Matlab and Simulink the client has opted not to commercialize this product at any level 5 3 Recommendations for Additional Work There are several features that were outside the scope of this project A wizard for adding new generator models is one such feature Another feature that could be created that would take significant work is an algorithmic parameter analyzer that would somewhat automate the process of machine parameter determination May07 18 Closure Materials Page 55 of 65 5 4 Lessons Learned This section details a few of the things the team has learned during the course of the last two semesters 5 4 1 What Went Well The team learned a great deal working on this project Things that went very well were the streamlining of the process of machine model entering once the prototype file structure was created During the second semester the team received a great deal of information from the client that allowed us to quickly and consis
34. ation 3 2 3 1 2 8 Zoom In Zoom Out and Pan Button Callback Functions The zoom in zoom out and pan callback functions are all very similar Each one changes the zoom property of each graph to either zoom in zoom out or pan disables the button that was pressed and enables the other two 3 2 3 2 GUI Setting File The GUI setting file is a Matlab script that creates three data structures The first parameter_setting contains the data necessary for the slider bars and edit boxes There is one entry in this data structure for each slider bar or edit box on the parameters panel The next output_setting contains all of the information necessary for the graphs on the output panel The last test_setting contains the information needed for the test selection menu These data structures are described in detail in Table 3 2 Three additional string variables simulin_file_name simulation_file and simulation_script are also placed in the workspace simulin_file_name contains the May07 18 Detailed Design Page 43 of 65 name of the Simulink block diagram simulation_file contains the title given to the home window simulation_script contains the name of the simulation script associated with that particular model An example setting file is included in Appendix B 3 2 4 Batch Simulation Window The batch simulation window shown in Figure 3 7 allows users to run several simulations at one time varying one parameter each time This is inten
35. can make the edit box and slider bar associated with certain parameters inactive based on the test selected in the drop down menu This is intended to reduce the possibility of entering a parameter in the wrong edit box A sub panel is also provided for each output variable Each sub panel has one plot box that consumes approximately 85 percent of the horizontal space and 92 percent of the vertical space The sub panels are arranged in a stacked manner with the horizontal x axis being much longer than the vertical y axis in any case where more than one graph is displayed 3 2 3 1 Structure A great deal of effort has been placed in to creating the entire graphical user interface as flexible and easily expandable as possible The home screen has been created in such a way so as to in theory allow an infinite number parameters and plots however due to space restrictions the user should attempt to limit the number of parameters to 96 The base code for the graphical user interface was created using GUIDE by building a basic window with the toolbar parameters panel and graph panel and then modifying the code to allow an infinite number of number of parameters and graphs Upon execution the parameter_setting model_setting and tests_setting data structures are collected from the Matlab workspace and are saved as global variables These are used throughout the program After the main window appears the interface waits for user commands After
36. ctions for the algebraic equations corresponding to the electrical network and loads It is very important that May07 18 Detailed Design Page 29 of 65 every integral have an initial condition defined It is not necessary to use Simulink models Instead a state space model could be formulated and one of the time domain differential equation solvers could be used to solve the system 3 1 1 4 Simulation Scripts The simulation scripts are Matlab m files that contain the equations necessary to calculate the initial conditions After all of the initial conditions have been calculated the Simulink model is executed 3 1 2 The Matlab Workspace The Matlab Workspace serves as a swap memory or information repository Any information that needs to be shared between subprocesses is sent to the base Matlab workspace and then read in by the next process This is intended to simplify function calling 3 1 3 Setting Files Setting files are Matlab scripts that create data structures in the Matlab workspace They are intended to store default data needed to properly initialize the window without requiring the user to enter it manually each time the program executes The main window setting file main window setting file m stores a list of the models in the library and the names of their respective setting files The several GUI setting files are specific to each model They contain the information necessary to initialize the home screen including th
37. ded to allow the user to see the effects of different values of one parameter and to aid in the overall parameter selection process When the batch simulate button is pressed the batch simulate window appears A pull down menu allows the user to select which parameter will be varied The maximum and minimum values correspond to the points between which the program will iterate By default these are set at the minimum and maximum slider bar values A callback function ensures that when the position of the pull down menu changes that the maximum and minimum values will change The simulations edit box allows the user to set the number of times that the program will run This number must be an integer greater than 0 When the OK button is pressed five variables are passed to the Matlab workspace parameter is an integer corresponding to the parameter selected in the pull down menu The maximum value max the minimum value min the number of simulations simulations and an exit flag cancel If cancel is set to 1 because the cancel button was pressed the home screen will not run the simulation if the OK button is pressed cancel will be set to O and the home screen will run the simulations 3 3 Simulation Scripts The simulation scripts are Matlab scripts that collect parameters from the workspace determine which test to run and initial conditions to use calculates additional initial conditions and executes the Simulink model Though these
38. duce different results Actual Result R2006a Same as R2006b R2006b Voltage Step Test Rcomp and Xcomp on Genrou model has no effect on output Vrmax on Genrou Exst4b has no effect on output Depending on other parameter settings simulation when Lpd is 1 sometimes results in simulation error Tr Vrmax Kf Tf Tf1 Tf2 Fbf Kii Vfmax Kh El Sel E2 Se2 on Genrou Rexs has no effect on output Rcomp on Gensal has no effect on output Vimax on Gensal ext has no effect on output Tb 0 simulation error Tr Vrmax Tf Llr Klr Velm Gmax Gmin on Gensal exstl hygov has no effect on output Action Required Simulation error cases are result of internal mathematics no changes necessary Verify with GE which parameters should be disabled for certain tests and include this information in setting file After discussion with GE it appears that they do not want these parameters disabled Completed Yes 2 1 5 1 2 6 Comparison of Simulation Data with PSLF Simulation Results Test Case Simulate all models with the parameters set at their default values and compare to PSLF data Purpose To verify that all models produce correct numerical results at their default values Procedure Use PSLF to prepare test data corresponding to the default case Export this data to a Microsoft Excel Spreadsheet whose format matches the one described in the detailed design Open each model from the opening window import the test data Press the
39. e disabled for each machine test and that a different event is simulated for each position Procedure Select each machine test simulate verify that the correct boxes are editable and that the simulation results are logical Expected Result Edit box are disabled properly a different test is simulated and the simulation results make sense for each test selected Actual Result R2006a Functions correctly see note R2006b Functions correctly see note May07 18 Approach and Results Page 23 of 65 Action Required note Disabled edit boxes need to be applied to exciters and governor parameters as discussed in Verification that Changing Parameters Produces Different Simulation Results section Completed Per a discussion with the client the edit boxes will not be disabled for exciter and governor parameters 2 1 5 2 2 10 Batch Simulations Test Case Batch simulation with the number of simulations set to a number less than 1 Purpose To verify that batch simulation has been properly implemented Procedure Click batch simulation button in lower left corner of home window choose a single parameter enter O into the simulations box and click OK Expected Result One simulation is run and the graphs appear properly with the correct legend Actual Result R2006a Same as R2006b R2006b Any number of simulations less and or equal to one is reset to two simulations Action Required none Completed Yes Test
40. e name of each parameter in the Matlab workspace the name that is to be displayed on the window the minimum value of each slider bar maximum and default values The GUI setting files also contain the workspace and display names for each of the values that are plotted and the names of each test or simulation case that the model can execute Each model must have its own setting file While creating setting files can be tedious this method does allow the user to add models to the library without modifying the base graphical user interface code or creating a newly coded graphical user interface 3 1 4 Test Data The program allows data to be plotted from Microsoft Excel spreadsheets This data must be saved in the xls format and shall be arranged according to Table 3 1 It is important that the first column correspond to time in seconds and the remaining columns correspond to the outputs There must be one and only one column for each output If no data is available then that column should be left as zeros but never blank The same format is used for exporting simulation data A sample of test data appears in Appendix A Table 3 1 Sample Test Data Format Time Output 1 Output 2 Output 3 Output 4 0 0 1 0 2 0 4 0 8 1 0 2 0 4 0 6 1 6 2 0 3 0 6 0 8 2 4 3 0 4 0 8 1 2 3 2 May07 18 Detailed Design Page 30 of 65 3 1 5 Interaction Between Components and Flow of Information The flow of information between sub processes is shown in detai
41. ect a generator type set generator parameters and view output waveforms Simulink models of each generator take the parameters from the GUI and calculate the output waveforms Other approaches that were considered included using the Java programming language to create the GUI Java is excellent for creating GUIs but does not interface with the Matlab environment very well Although the Matlab language is not as well set up to design complex GUIS it interfaces with the Simulink models more efficiently In the end it was decided to use Matlab because the additional features available in the Java interface did not provide any benefits in this context An end user manual test plan and this final report are provided to ensure that the client fully understands how to use the product how it was tested for robustness and add additional machine models to the interface in the future 1 4 Operating Environment The product was designed using Simulink and Matlab2006a Matlab2006b it will be run on Windows XP based computers with monitor resolution greater than 1024x768 Matlab and Simulink can be run on other operating systems and screen resolutions but has only been tested in this environment May07 18 Introduction Page 2 of 65 1 5 Intended Users and Uses This section defines the intended users and uses of the project team s solution 1 5 1 Intended User s This software will be used by exclusively by employees of General Electric It is a
42. equired none Completed Yes 2 1 5 2 2 11 Save Output Plots as jpg Files Test Case Save output plots as JPEG images Purpose To verify that the save output plot feature has been correctly implement Procedure Run a simulation click the save button and save the output plots as test jpg Look in the directory to see if the correct number of figure files are there and that the filenames are correct Use windows to open the figures and compare them to the output plots Expected Result The correct number of figures will be created with the filenames as described in the detailed design and the figures will match the output graphs Actual Result R2006a Same as R2006b R2006b Jpg saves correctly however legend is missing Action Required Add legend to jpg captures Completed Yes 2 1 5 2 2 12 Graph Previous Simulation Data Test Case Graph previous simulation data Purpose To verify that the previous simulation data is graphed correctly May07 18 Approach and Results Page 25 of 65 Procedure Run a simulation change the event time from the default to 5 seconds press the simulate button Check to see that the previous simulation data is plotted with the correct event time as a light gray line and that the current data appears as a colored line with the same numerical results at a time delay Verify that the correct legend appears Expected Result The previous simulation results should appear as a light gray line with
43. er panel test selection pulldown menu of the home window Found in the GUI setting file 3 2 2 Opening Window The opening window is shown in Figure 3 4 and serves as a user friendly interface for model selection Upon execution the main window setting file main_window_setting_file m executes placing the model_setting data structure into the Matlab workspace The model_setting data structure contains the names of each model and the name of the corresponding GUI setting file After this executes a screen similar to Figure 3 4 appears May07 18 Detailed Design Page 34 of 65 lala Choose machine model Figure 3 4 Main Window The program waits for the user to select a model and press the Open button After the Open button is pressed the program executes the GUI setting file that corresponds to the selected model executes placing the parameter setting output_setting and tests_setting data structures into the workspace The GUI setting file is described in more detail in Section 3 2 3 2 3 2 2 1 Structure The base code for the main window was generated by the Matlab GUIDE toolbox The create function contains code that executes the main window setting file placing the model_setting structure in the Matlab workspace This data is then read into a global variable that is used in the layout and Open button callback functions The callback function associated with the Open button was modified t
44. erify zoom out feature is implemented properly Procedure Run simulation with out test data loaded zoom out on the graphs by pressing the zoom out button and using the left mouse click Zoom in on graphs May07 18 Approach and Results Page 21 of 65 using the shift key plus left click simulate one more time after zooming in and out to ensure that the graphs hold the current zoom level Expected Result Graphs zoom in and out properly graphs reset after each simulation Actual Result R2006a Works correctly R2006b Works correctly Action Required none Completed Yes Test Case Pan plots without test data Purpose To verify that the pan feature is properly implemented Procedure Run simulation without test data loaded pan graphs in different directions after zooming in simulate afterward to ensure that graphs reset Expected Result Graphs pan properly and reset after simulation Actual Result R2006a Works correctly R2006b Works correctly Action Required none Completed Yes Test Case Pan plots with test data Purpose To verify that the pan feature is properly implemented Procedure Run simulation with test data loaded pan graphs in different directions after zooming in simulate afterward to ensure that graphs do not reset Expected Result Graphs pan properly and hold current position after simulation Actual Result R2006a Works correctly R2006b Works correctly Action Required none Com
45. es perform as intended For each functional requirement there will be a set of test cases and corresponding expected results for each feature associated with that requirement The actual test results are then compared with the expected results and documented in 2 1 5 2 2 for future reference 2 1 5 2 1 1 Items Included in Functional Testing The following is a list of functional requirements included in functional testing Start exit the program from Matlab Select a machine model to simulate from main window Import test data and export simulation data Save and open simulation data Edit parameter values using the edit boxes and slider bars Undo redo changes using undo and redo buttons Plot simulation results Zoom and pan plots May07 18 Approach and Results Page 13 of 65 e Overall appearance for different screen sizes e Switching between different machine tests e Batch simulations e Save output graphs as Matlab figures e Graph previous simulation data e Time required to determine the parameters should be less than four hours This will be completed by the team at General Electric 2 1 5 2 1 2 Items Excluded from Functional Testing Here is the list of requirements excluded in functional testing e Non functional requirements such as ease of use e Setting files are tested under model testing and are assumed to be properly configured 2 1 5 2 2 Test Cases Here is the list of test cases and expected results for each
46. etermining what they really needed This may have alleviated some confusion about program functionality and aesthetics and would have eliminated some of the last minute feature additions 5 5 Risk and Risk Management This section lists anticipated and unanticipated risks and some of the mitigation techniques used to minimize the risk May07 18 Closure Materials Page 56 of 65 5 5 1 Anticipated potential risks and planned management The fact that this is a software project alleviated many of the common concerns regarding ordering of parts and what to do if expensive parts were damaged or important parts were not available The biggest risk management issue that the team faced was eliminating the possibility loosing important data Because our entire project is a few files it was very important that the team back them up in multiple places The group set up an FTP server that was used not only to share data amongst group members but also to ensure that multiple copies of all files were maintained This server was backed up several times a week 5 5 2 Anticipated risks encountered and success in management Fortunately data loss did not pose a problem at any point during the project and no files were ever accidentally deleted or lost 5 5 3 Unanticipated risks encountered attempts to manage and success An unanticipated risk that the group encountered was the release of a new version of Matlab while the project team was still develop
47. f synchronous generators are essential to asses the stability of a power system The current process used by General Electric to determine machine model parameters consists of running tests under known conditions in the field and recording extensive amounts of data related to machine performance After returning to the office General Electric engineers estimate model parameters using the engineer s own intuition and experience load these into PSLF and run simulations of the field tests The simulation results are then compared to the results of the actual field tests The engineer iteratively matches the test data to the simulation data until an accurate parameter setting for the generator is determined Although the data collection for field tests only take a few hours the process of determining accurate model parameters takes an experienced engineer between three and five days with currently used PSLF software The client desires a well documented interactive tool designed using Simulink and Matlab that will reduce the time required to determine generator parameters to less than 4 hours 1 3 2 General Solution Approach The project team met with the client and faculty advisor to develop a more complete understanding of the project and create a prototype block diagram and graphical user interface Various approaches to both the design of the GUI and machine models were considered The end design uses a Matlab based GUI that allows the user to sel
48. functional requirement listed in 2 1 5 2 1 1 The test case defines what will be tested The reason for testing this is described in the purpose Procedure outlines how the test will be performed The expected result explains what result will be considered completely successful The actual result describes what happened when the test was executed and what modifications were necessary to achieve the expected result If the expected result was not achieved and no attempt to remedy the situation was made a justification should be given 2 1 5 2 2 1 Start and Exit Program From Matlab Test Case Start program from Matlab Purpose To verify user can start the program correctly by following the instructions outlined in the user manual Procedure Have testers without knowledge of the software follow the procedure outlined in the user manual to start the program inside Matlab Expected Result Tester should be able to start the program within Matlab and opening window should appear correctly R2006a Same as R2006b R2006b Issues found see below Action Required Highlight circle current directory in Fig 2 1 of user manual Completed Yes May07 18 Approach and Results Page 14 of 65 Test Case Exit opening window Purpose To verify user can exit opening window using the close button and that the opening window closes properly Procedure Close opening window using the close button in the upper right hand corner Expected Result
49. hite white white white white white white white ssa white white Ub ole P white white white l 44 4 t 4 41 1 4 41 1 4 4 1 4 41 11 4 2 4 1 4 4 ced uw x gelu mr di Freia slider_panel Lg 1 1 1 1 1 14s Lrl1 1 1 01 1 T 22 2 2 2 2 24 Ay 2 2 2a een 2 252 3 3 3 3 3 3 3 4 4 yan editability 0 0 0 0 0 0 0 0 All Tests 0 0 0 0 0 0 0 0 1 Test 1 Only LEUR PR In 2 Test 2 Only 24252 3 Test 3 Only 34 3 3 4 Test 4 Only Ty 1 etc May07 18 Sample GUI Setting File Page 62 of 65 NO OW p07 O 07 Oy oss OoOWN OO WwW output setting struct workspace name uit Efd spd ledifd ads start time Ob bro end time 10 display name Vi t pu Efd pul speed pu ladifd pu ae test data Vs Ejere simulation data x k old simulation data x l test time 0 tests setting struct test_name Voltage Step Load Rejection Three Phase Fault ETT simulin_file_name SMIBgensal_v1 simulation_file Gensal Model simulation_script SMIBgensal_sim May07 18 Sample GUI Setting File Page 63 of 65
50. ials 75 Total 6750 Table 4 4 Financial Costs Associated with Resources Actual Costs Resource Cost Labor 12 010 Materials 104 Total 12 114 4 2 Schedules This section contains the spring and fall schedules showing original estimates revised estimates and actual times Figure 4 1 shows a legend for the spring and fall charts Figure 4 2 shows the fall 2006 semester schedule Figure 4 3 shows the spring 2007 schedule and Figure 4 4 shows the deliverable dates Table 4 5 shows task dates for the original schedule and Table 4 6 shows the final task dates The only significant differences between the original and final schedules are the lengths associated with design and implementation of the prototype and end product The deliverables dates have not been changed so there is only one schedule associated with it May07 18 Resources and Schedules Page 50 of 65 Figure 4 1 GANTT Chart Legend ID Task Name 7 06 Sep10 06 Sep 24 06 Oct8 05 Oct 22 06__ Nov 5 06 Nov19 06 Dec 3 06 Dec17 06 SITITISIFIWIM S ITITISIF WIMISIJ JTITISI F WIMISIT T 1 Task 1 Project Definition 5 Task2 Technology and implementation considerations an 10 Task 3 Prototype Design 14 Task 4 Prototype Implementation 17 Task 5 Prototype Testing 23 Task 6 Feedback
51. imulation data should appear as a light gray line to allow the user to judge progress in matching the test data Color Code for Edit Box Supplemental Color edit boxes to help user to visually group different types of parameters belonging to different subsystems 2 1 2 Design Constraints The constraints outlined below must be considered e The program must run completely within the Matlab R2006a R2006b program environment including the machine model running in the Simulink environment e If the parameters are varied the output displays must be able to update in semi real time The results must appear within a few moments May07 18 Approach and Results Page 6 of 65 The complexity of the block diagram and number of parameters is generator specific The GUI for each generator will be tailored to its specific parameters and outputs while maintaining standardization 2 1 3 Technical Approach Considerations The technical aspect of the project was approached in the following way Full generator models were provided by the client for the team to integrate into the software package It was determined that the team did not have the technical expertise to create models of General Electric s generators from scratch in the amount of time provided Although there is some supplementary Matlab text based code the generator models were created using Simulink in graphical block diagram form so that the structure is apparent The gra
52. in the field that is compared to simulation results Simulink and Matlab are copyrighted by The Mathworks Inc Windows is copyrighted by Microsoft Corporation PSLF is copyrighted by General Electric Company May07 18 Introduction Page vi of vi 1 Introduction This section defines the purpose and major objectives of the project including a project description discussion of operating environment intended users and uses assumptions and limitations and expected end product and deliverables 1 1 Executive Summary In order to analyze the security and stability of a power network engineers use software such as General Electric s PSLF to test their generators under various conditions In order for these simulations to be accurate detailed models of synchronous generators are needed These models have a great number of parameters that are determined by testing the machines under known conditions in the field and varying simulated parameters based on the engineer s intuition and experience until the model produces the same results as the field test Although the field tests only take a few hours the process of determining generator model parameters can take an experienced engineer between three and five days Software was needed to help engineers determine these parameters in less than a day 1 1 1 Project Activities The project was divided up into 4 main phases The first phase consisted of research on generators and collecting inf
53. ing the end product Unfortunately this caused problems with virtually every component of the program The client wanted a project that would be compatible with both the new and previous version and the project team struggled to find methods of zooming and panning the plots that would work on both versions Solutions were found that allowed all of the features to work on both versions 5 5 4 Resultant Changes in Risk Management Made Because of Encountered Unanticipated Risks Because of the need for the project to work well with both versions of Matlab all releases are tested on both versions to ensure compatibility 5 6 Project Team Information Jared Kline Electrical Engineering German 422 Stonehaven Dr 13 Ames IA 50015 Phone 515 480 2606 Email jkline iastate edu Mark Reisinger Electrical Engineering 1831 Fuller Road 18 West Des Moines IA 50265 Phone 319 621 4986 Email bluemonk O iastate edu May07 18 Closure Materials Page 57 of 65 Adam Wroblaski Electrical Engineering 1106 Pinon Dr 5 Ames IA 50014 Phone 515 441 1396 Email youareme iastate edu Yu Chan Computer Engineering Mathematics 315 Welch Ave Ames IA 50014 Phone 515 292 5171 ext 211 Email ychan iastate edu 5 7 Client Information General Electric Company Primary contact Doug Welsh Manager Software Products Address 1 River road Bldg 2 Rm 623 Schenectady NY 12345 Telephone 518 385 5619 Mobile 518 281 4341 Em
54. l in Figure 3 3 Entry Point os Execute Opening Execute GUI Setting File Command Window Command Setting File Opening Opening indow ome Window Data Window Data Data Window Setting Simulation Results Simulation Batch Simulation Data Results Home Matlab Screen Workspace rerum Parameters Parameters amp Batch Initial Conditions Execute Simulation Kommand Data Initial Conditions Parameters Execute Batch Execute Command Simulation Command Window Simulation Script Figure 3 3 Detailed Flow of Information As one can see the Matlab workspace serves as the central information storage point All data that is shared among processes is stored in the Matlab workspace 3 2 Graphical User Interface The graphical user interface can be broken into three components the opening window main_window m home window GUl m and the batch simulation window batch_simulation m Each of these components is described in detail in sections May07 18 Detailed Design Page 31 of 65 3 2 2 3 2 4 A discussion on the data structures that store all of the information necessary appears in Section 3 2 1 3 2 1 Data Structures Several data structures are used throughout the program a summary of these appears in Table 3 2 The model setting data structure is used by the opening window There is one data structure for the entire library The parameter setting data structure contains the information needed for the parameters
55. lation data so that the gray trace associated with the previous simulation will not appear when the simulation button is next pressed Each parameter is then checked to see if its status should be changed from active to inactive or vice versa 3 2 3 1 2 3 Simulate Button Callback Function The simulate button callback function first makes the simulate button inactive and changes the text from Simulate to Simulating Next a wait bar is created and each parameter is placed in the workspace If this test has been simulated previously the old simulation data is placed into the old simulation data field in the output setting data structure and the simulation is executed If there is a problem with the simulation a catch statement will cause a message box displaying Simulation error to appear This is shown in Figure 3 7 This prevents the program from crashing but does not provide much insight into the source of the problem If the simulation executes properly the results are plotted and a legend appears in the upper right hand corner of the top plot box A custom color scheme is defined to ensure that all of the plots are legible and that all colors are pleasing to the eye olx ES Simulation error Figure 3 7 Simulation Error Message Box 3 2 3 1 2 4 Batch Simulate Button Callback Function The batch simulate call back function starts by making the batch simulate button inactive and changing the text from Batch Simulate
56. m level should not change after the simulate button is pressed Graph Previous Simulation Supplemental Data The previous simulation data should appear as a light gray line to allow the user to judge progress in matching the test data Color Code for Edit Box Supplemental Color edit boxes to help user to visually group different types of parameters belonging to different subsystems 3 2 3 1 Callback Functions Callback functions are used to specify what actions should occur after a certain action is performed The callback functions necessary for the completion of the main window appear in Sections 3 2 3 1 2 1 through 3 2 3 1 2 8 3 2 3 1 2 1 Slider Bar and Edit Box Callback Functions The slider bars share a common callback function The edit boxes do as well This is possible because each slider bar and edit box has its own handle a double identification number that is unique to that object The remaining buttons have their own callback functions The edit box callback function collects the new value from the May07 18 Detailed Design Page 39 of 65 edit box and changes the position of the slider bar The slider bar s range is adjusted if necessary The slider bar call back function contains code that updates the value in the edit box and adjusts the position of the slider bar 3 2 3 1 2 2 Test Selection Menu Callback Function The test selection drop down menu call back function first deletes both the current and old simu
57. n engineering to depict the relationship between inputs and outputs of a system Exciter System that generates the rotor currents needed to produce electric flux Genrou Round rotor General Electric generator model Gensal Salient pole General Electric generator model GUIDE Graphical User Interface Development Environment A Matlab toolbox that supplies a graphical and interactive method for the creation of graphical user interfaces Initial Conditions Values that determine the simulation s starting point Matlab A software package developed by Mathworks that is commonly used for engineering computation Matlab Workspace The location where all variables are stored Physical test results These are the results of actual generator performance as physical measured in the field PSLF Positive Sequence Load Flow A software tool manufactured by General Electric that is used by power systems engineers to analyze the performance and security of large interconnected power systems PSLF test results These are the results of control simulations run using PSLF Semi real time Result updates appear after the user selects the run command Setting File GUI reads setting files to customize it s appearance Simulation Results Output data from the Simulink based generator models Simulink A Matlab toolbox that allows users to build and analyze block diagrams Test results Data collected
58. nal Labor Estimates ei t eret o lessened sussadevaneddeds seceans esee 46 Table 4 2 Actual Eabor Usal nsnsi en teeta eek Atlee eel eats 48 Table 4 3 Financial Costs Associated with Resources Original Estimate 50 Table 4 4 Financial Costs Associated with Resources Actual Costs 50 Table 5 1 Project Task Byaluauon ederent test i 53 Table 5 2 Nhlestone Evaluation TA Gees 55 May07 18 Page iv of vi List of Figures Figure 3 1 Simplified Flow of Information ccsccesseeesseeseeceneeesoneescetseecenseesonteses 28 Figure 3 2 Simplified Plow Chart cin repe edt tet eene vade Geen tacts 29 Figure 3 3 Detailed Flow of Information uique eds nde etae oet i ciis ndi cbssbdedeted un 3l Figure 3 4 Mam Window ies iate odveiis tcs toten ip todas ob secs itii diia 35 Fig re3 gt Home WIN A 36 Eigure 36 Wait Daten teer u od a iie 38 Figure 3 7 Simulation Error Message BOX sees 40 Figure 3 8 Batch Simulate Wind Wii a 41 Ergure 39 Open WIHOO enis iio eH a dal ie alee sede REA 42 Figure 3 10 Saye As WIBdOW iiid tct cun Siete bd 43 Figure 4 L GANTT C hart Legend edente dc dee 51 Figure 4 2 Fall2006 S hed le o dea b Cee 51 Figure 4 5 Spring 2007 Schedule eec ee E E a 51 Figure 44 Deliverables Schedule adn eto e pott cessit oct cni armen de 52 May07 18 Page v of vi List of Definitions Block diagram A visual method commonly used i
59. nvalid test data Purpose To verify home window still functions properly after attempting to import improperly formatted test data Procedure Prepare test data formatted in a format other than the format described in the detailed design and attempt to import it Expected Result Exception message is displayed and the program still functions properly Actual Result R2006a Same as R2006b R2006b When invalid numbers were put into time column or plot data column no errors occurred just gaps in plot When columns were in wrong location a detailed error message was given Action Required none Completed Yes Test Case Export simulation data after simulation Purpose To verify export simulation data feature has been correctly implemented Procedure Click save button and verify that exported simulation data matches the simulation results Expected Result Exported data matches simulation results displayed on the home screen Actual Result R2006a Same as R2006b R2006b Jpg saves correctly however legend is missing Action Required Add legend to jpg captures Completed Yes May07 18 Approach and Results Page 16 of 65 Test Case Export simulation data before simulation Purpose To verify program still functions properly after attempting to export simulation data when none exists Procedure Click export button before running any simulations when home window starts Expected Result Exception message ap
60. o execute the simulation script corresponding to the selected model The main window can support any number of models 3 2 2 Main Window Setting File The main window setting file is a script file that creates a data structure named model setting and places it in the Matlab workspace This data structure contains two fields for each entry one for the model s name model name and the other for the name of the GUI setting file that corresponds to the model May07 18 Detailed Design Page 35 of 65 3 2 3 Home Window ini xj GEoS5290 r Parameters ladifd 177 T T T T T Voltage Step z Simulation m Tpdo Tppdo Tppqo Jew r THAT r Lpp 1 755 Lpd pus poz EN spd S 10 15 20 25 30 mU JJ 1 Ex Ra r Rcomp Xcomp amp a jump x 7 Pgen_Ystp Qgen_Vstp W1_Vstp mu E 100 3208 kx 105 i Pgen_Rict Qgen Rjct V1 Rit 0 5 10 15 20 25 30 zl 100 32 09 105 Eta m Pgen FI Ggen Fit vi FE 2 dE qs Tup Vstep 36V Vstep T Rit E 1 a 1 a 1 T_Mt_on T fit off Ztautt 1 1 0 4755 1 1 L L L Rt xt RI Jm m m o X MVA mach MVA sys mm um fbase Tstop Step Size Jia as
61. on all blocks in the algebraic equations in order to maximize flexibility A sample Simulink block diagram appears in Appendix C May07 18 Detailed Design Page 45 of 65 4 Resources and Schedules This section gives a detailed schedule that was followed by the team throughout the project It also contains the resources that were used during the course of the project It includes the original schedule and resource estimates predicted for the project as well as the actual usage for the course of the project 4 1 Resource Requirements The resources that were required for the project completion were mainly time and expertise provided by the team members advisor and the client Other resources such as computer hardware and software were provided by the University with the exception of poster board lamination document binding and shipping 4 1 1 Personnel Effort Requirements This section shows the chart representing man hours based on our Gantt chart schedule Table 4 1 shows our original estimates and Table 4 2 shows the actual usage for the entire project The only major differences between the original estimates and the actual are the allocations for design and implementation We have adjusted the numbers showing that implementation took more time than originally estimated Table 4 1 Original Labor Estimates Task Mark Adam Jared Yu Task Total Task 1 Problem Definition 4 4 2 4 14 Task la Problem Definition Completion 4 2 0
62. ormation from the client The second phase consisted of planning the design process The third and longest phase was an iterative process of prototyping and feedback from the client Followed by fine tuning bug testing and documentation in the fourth and final phase 1 1 2 Final Results The project team created a software package that met or exceeded the expectations of all required and supplemental specifications It allows engineers to determine the parameter settings on their generators much more efficiently The team feels that the project is a tremendous success and will save the company time and money 1 1 3 Recommendations for Additional Work There are several features that were outside the scope of this project A wizard for adding new generator models is one such feature Another feature that could be created that would take significant work is an algorithmic parameter analyzer that would somewhat automate the process of machine parameter determination 1 2 Acknowledgements The project team would like to thank Doug Welsh Dan Leonard and Juan Sanchez of General Electric and Dr Chen Ching Liu of Iowa State University for the assistance and technical expertise that they have provided May07 18 Introduction Page 1 of 65 1 3 Problem Statement The general project statement and general project approach explain the problem and the approach that the team used for the solution 1 3 1 General Problem Statement Accurate models o
63. panel of the home screen There is one parameter setting structure for each model The output setting structure contains the information necessary for the output panel of the home screen There is one output setting for each model The tests setting structure is again model specific and contains the information necessary for the test selection menu on the home screen May07 18 Detailed Design Page 32 of 65 model_setting parameter_setting Table 3 2 Data Structures and their Fields Stores the data needed by model_name the opening window the name of each model in the library and the setting file associated with that model There is one structure for the entire library and one entry setting file name for each model Found in main window setting file Stores the data that is workspace name needed for the parameters panel on the home window This information is also used by the batch simulation window There is one min value structure for each model and one entry for each Found in GUI max value default value previous value current value display name String value of the mode s name as it should be displayed on the home window String value of the setting file associated with that particular model String containing the name of the parameter as it will appear in the workspace and will be used in the simulation script and Simulink model parameter value parameter value paramete
64. pears and program still functions properly Actual Result R2006a Same as R2006b R2006b Jpg saves empty plots non issue xls does not save no error message given Action Required Add exception to display message notifying user that xls was not saved Completed Yes 2 1 5 2 2 4 Save and Open Parameter Data Test Case Save parameter data Purpose To verify that the save parameter data function has been correctly implemented Procedure Use the save button to save the data as a Matlab data structure Use Matlab to view the data structure Check to see if the data in the structure is correct Expected Result Data will be saved correctly Actual Result R2006a Same as R2006b R2006b Parameter data saves correctly Action Required none Completed Yes Test Case Open previously saved parameter data Purpose To that the user can open saved parameter data and that the edit boxes and slider bars are set to the correct position May07 18 Approach and Results Page 17 of 65 Procedure Use the open button to import previously saved parameter data Check to see that the edit boxes now display the correct values Check that the slider bars are in the correct position Expected Result User can import previously saved data and that data will be displayed properly on the home window Actual Result R2006a Same as R2006b R2006b Parameter data opens correctly Action Required none Completed Yes Test Case
65. phical user interface window contains everything that the user needs to easily vary the parameters and analyze the outputs of the machine It was determined that having all of the parameter values and outputs plots in one window would provide a faster and simpler interface for the engineers Having multiple windows was considered but determined to be less efficient The software package runs entirely from the Matlab program It was determined that although other languages such as Java could offer more flexibility in the design the Matlab graphical environment would perform the required functions as well as interacting with the Simulink files in a more efficient way The Matlab workspace is the holding place for all variables and other data passed between the graphical user interface and Simulink models A direct link was considered between the GUI and models This approach was not chosen because the intermediary space is easily accessible by both which allowed for more efficient software coding Also in the event of a simulation or GUI error the current data will not be lost Output graphs are of the same form as those generated by testing and are displayed in the GUI in such a way that they can be easily compared by the user It was determined that the graphical user interface should be able to handle different time step values for the simulation results to match the precision of test results from the field The parameter slider bars and ent
66. pleted Yes 2 1 5 2 2 8 Overall Appearance for Different Screen Sizes Test Case Slider bar and edit box appearance at large and small screen size May07 18 Approach and Results Page 22 of 65 Purpose To verify slider bars and edit boxes appear and resize correctly Procedure Start home window with various numbers of parameters in setting file and resize window by clicking and dragging the bottom right corner Expected Result All slider bars and buttons show up completely and are clickable text is readable in edit box and is shown to at least 4 significant digits Labels are also readable Actual Result R2006a Same as R2006b R2006b Appearance is sufficient at minimum of 1024x768 pixels Action Required none Completed Yes Test Case Graph appearance at large and small screen size Purpose Verify that graphs appear correctly Procedure Run program with setting files that have different numbers of output graphs to determine the maximum feasible number of output plot boxes Expected Result All plots labels and axes appear correctly up to a maximum value Actual Result R2006a Same as R2006b R2006b Appearance is sufficient at minimum of 1024x768 pixels Smaller windows still function however there are some readability problems Action Required none Completed Yes 2 1 5 2 2 9 Switching Between Different Machine Tests Test Case Select machine test from drop box Purpose Verify that correct edit box ar
67. r value Double containing the previous parameter value Used for undo redo functionality parameters current value String containing the name of the parameter as it will appear on the parameters panel A three double array that determines the color of the edit box An integer that determines for which tests the parameter is active If this is set to 0 the parameter is always editable if itis set to a number greater than 0 it will only be editable if the test pulldown menu is set to that position May07 18 Detailed Design Page 33 of 65 Table 3 2 Continued output_setting Stores the data that is needed workspace name String containing the name of for the ouput panel on the the variable as it appears in the home window There is one Simulink block diagram and structure for each model and workspace one entry for each output Not used on current versions Found in the GUI setting file Not used on current versions is displayed on the graph s Y test data Double array containing the data imported from the Excel spreadsheets simulation_data Double array containing the simulation data old_simulation_data Double array containing the data from the previous simulation test_time Double array containing the time from the imported Excel data tests_setting Stores the strings that are test_name String containing the name of needed for the test selection the test as it is displayed in the menu on the paramet
68. rocess May07 18 Introduction Page 4 of 65 2 Approach and Results This section outlines the design team s solution to the client s problem steps that were taken for the design to be successful as well as the functional requirements for the final product 2 1 Approach The project is deemed successful if it fulfills the following requirements 2 1 1 Functional Requirements The required and supplemental functionalities are listed in Table 2 1 Table 2 1 Product Requirements and Features Required or Supplemental Intended Functionality Required Provide the user with a means to quickly enter a specific value for a parameter Required Provide the user with an easy method to vary parameters Each slider bar shall have a default maximum and minimum value When the slider bar position is changed the value in the edit box shall update If a value is entered into the edit box that is outside of this range the corresponding maximum or minimum should reset to the value in the edit box Interface with Simulink Required The user shall be able to execute Simulink models from the graphical user interface Graph Simulation Data Required Simulation data shall be plotted on figures in the graphical user interface Import Data from Excel and Required The user shall be able to import time Display on Graphs domain data stored in Microsoft Excel spreadsheets in a format specified by the project team and plot them on the same
69. rogram attempts to open it as a mat save case If this succeeds each parameter is set to the value in the file if this fails as well an error message is displayed 3 2 3 1 2 6 Save Button Callback Function The save button callback function starts by opening the standard Windows Save As dialog box This is shown in Figure 3 10 The user has the option of saving three different files The first is to save the output plots as Matlab figures fig If this option is chosen the program will save each graph individually with the workspace name of the output variable concatenated to the end of the filename If the user elects to save the parameter data save case as a mat file the parameter_setting array is saved If the user wishes to save the simulation data the simulation data is saved to an Excel spreadsheet in the format in Section 3 1 4 May07 18 Detailed Design Page 42 of 65 Save As ax Save in C3 gui_03 08 2007 E3 My Documents rr My Computer Figure 3 10 Save As Window 3 2 3 1 2 7 Undo Redo Button Callback Functions If the Undo toolbar button is pressed the program changes the last modified parameter to its previous value and enables the redo button If the redo button is pressed the last undone parameter is changed to its pre undo value It should be noted that the undo functionality is only available after the simulate button has been pressed and can only undo changes up to the previous simul
70. roperly according to the sliderbar position Expected Result Value inside edit box updates properly according to the sliderbar position Actual Result R2006a Sliders act correctly R2006b Sliders act correctly Action Required none Completed Yes 2 1 5 2 2 6 Undo and Redo Changes Using Undo and Redo Button Test Case Undo changes using undo button Purpose To verify that undo button has been properly implemented Procedure Press the simulate button change 20 parameters noting previous and current values press the simulate button then click the undo button 20 times Verify that the values are reset to the correct value Expected Result Parameter values change back to previous values each time undo button is pressed Actual Result Actual Result R2006a Undo button performs correctly R2006b Undo button performs correctly Action Required Identify undo redo disabling code and check for uniformity Completed Yes Test Case Redo changes using redo button Purpose To verify redo button is implemented properly Procedure Press the simulate button change 20 parameters noting previous and current values press the simulate button then click the undo button 20 times Verify that the values are reset to the correct value then click the redo button 20 times verify that the values are set to the correct values Expected Result Parameter values change back to previous values each time redo button is pressed Actual
71. rting A a 3 o s Tast a Project Plan Development o A Task 120 Design Repor Development a o AAA Tsk Te Proje Poser eran 5 3 TA a LIT SUI eee Pee OOOO OOO AE E IE Meetings and Correspondence 90 Grand Total 351 1201 4 1 2 Other Resource Requirements Computers running Microsoft Office Matlab 2006 and Simulink were required for both programming and documentation through out the project Server space was also required for sharing files between team members and hosting the project website Our original estimate for binding was 25 this was changed to 72 because there were 4 documents being bound at 13 each and 5 shipping to the client per document Poster printing was done at no charge by the department so our original estimate of 50 was reduced to actually 12 for lamination and 20 for poster board and glue Computers required were available in the computer labs provided by the University The 104 binding and poster fee was split between team members May07 18 Resources and Schedules Page 49 of 65 4 1 3 Financial Requirements This section details the cost associated with the resource requirements The original estimate of project costs appears in Table 4 3 the revised estimate appears in Table 4 4 the reason for the increased cost is the increase in estimated labor Table 4 3 Financial Costs Associated with Resources Original Estimate Resource Cost Labor 6670 Mater
72. ry boxes are color coded to represent their function in the machine and separate them visually to the user Testing revealed that separating the parameters boxes with white space and headings took too much space away from the output plots when used in a low monitor resolution The client specified that the large size of the plots was more important for accuracy Setting files have been created for each machine detailing which parameters will be in use for all available tests Parameters edit boxes that are in use for a May07 18 Approach and Results Page 7 of 65 machine but not for the selected test are disabled It was determined the GUI should only display the parameters edit boxes that apply to that machine in order to save space for the output plots It was also determined that removing and replacing parameter edit boxes for each individual test within a machine would slow down the user experience too much 2 1 4 Implementation Process Description After developing a robust and usable prototype the primary implementation activities for the project consisted of obtaining models settings and parameters for all of the different generators that were going to be used in the program A significant hurdle was overcome early in the implementation process through the development of setting files and parameter files that interface with the main GUI This allowed the team to rapidly develop the necessary configuration settings for each model witho
73. sic components the graphical user interface the Simulink models and the simulation scripts that connect them The Matlab workspace serves as a common location for stored data In virtually every location where one subprocess interfaces with another variables are stored in the Matlab workspace by one process and called by another This greatly simplifies the calling of functions as variables do not need to be explicitly passed The simplified flow of information between the basic program components is shown in Figure 3 1 May07 18 Detailed Design Page 27 of 65 Simulation Results Simulation Results Matlab Workspace Simulink Parameters Parameters amp Initial Conditions xecute Command Execute Command Simulation Script Figure 3 1 Simplified Flow of Information 3 1 1 1 Basic Algorithm The basic algorithm is shown in Figure 3 2 After the initial program execution the opening window appears The user selects a model and presses OK The home screen appears and the user imports test data and chooses to either batch simulate run several simulations at once with slight adjustments to one parameter or to edit parameters himself and simulate The results are plotted with the test data and the user compares the results If they are satisfactory he exits the program if they are not he can choose to edit parameters or run more batch simulations A more detailed version of this flowchart appears in Appendix D
74. ssumed that they are experienced engineers who are familiar with Simulink and Matlab generator stability and machine testing 1 5 2 Intended Use s This software is used to aid engineers in quickly determining model parameters for synchronous machines 1 6 Assumptions and Limitations This section defines the initial assumptions and limitations that the group worked from to solve the problem 1 6 1 Initial Assumptions Table 1 1 lists the initial assumptions that the group worked from to find and implement the solution to the problem Table 1 1 Initial Assumptions and Justifications Assumptions Justification The project team will be provided with models of all Implementation relevant machines to be included before project completion The project team will have access to the results of Testing relevant PSLF simulations for results comparison Project team shall not share information supplied by Confidentiality client or design results with outside individuals or groups Client s version of Matlab will be compatible with Design project team s The program interact with the user in a graphical way Design 1 6 2 Initial Limitations Table 1 2 lists the initial limitations with which the group will comply May07 18 Introduction Page 3 of 65 Table 1 2 Initial Limitations and Justifications Limitations Justification The software will be created using Matlab and Simulink Client
75. tently add features and root out bugs in the system 5 4 2 What Did Not Go Well The things that did not go well were our inability to create machine models on our own in the beginning of the project Developing the models from scratch is normally considered advanced graduate material and was simply outside the realm of the project team s understanding 5 4 3 What Technical Knowledge Was Gained The team gained a great deal of knowledge about the functions of Matlab and how to create programs in the environment The team also learned how to use different parts of Matlab to pass information back and forth for data analysis The project team also gained a great deal of understanding of the considerations of machine operation testing and modeling Power systems engineers and professors were consistently very impressed with both the difficulty of the project and the depth of understanding that team members had for complex and difficult material 5 4 4 What Non Technical Knowledge Was Gained Team relationships proved to be a major struggle for our team but offered a powerful learning environment for interpersonal relationships in a team oriented technical project 5 4 5 What The Team Would Do Differently If The Team Had To Do It Over Again If given the opportunity to start again with the knowledge the team have now the team would have spent more time in the beginning of the project gathering requirements interfacing with the client and d
76. the default event time The current simulation data should appear as a colored line with the same magnitudes as the previous simulation but with a time delay The legend should be correct Actual Result R2006a Works correctly R2006b Works correctly Action Required none Completed Yes 2 1 5 2 2 13 Time Required to Determine Machine Parameters Test Case Test the speed with which a set of parameters can be found using software package Purpose To verify that the time requirement is met Procedure Have a member of the General Electric team determine machine parameters using test data from an unknown generation facility Time the engineer Expected Result Parameters will be determined in less than 4 hours Actual Result Preliminary results done by GE indicate expected result is on target more testing will be needed to confirm 2 2 Project End Result The final outcome of the project was a complete success The client was able to accurately match the programs simulation values to a more cumbersome but very accurate simulation program that has been in place with the company Additionally the client was able to take fresh data from the field and make use of this program to analyze the parameters of the generator On the team s side everyone felt as though the experience was very helpful because of the significant Matlab and Simulink knowledge learned will surely help in the future May07 18 Approach and Results Page 2
77. the simulate button is pressed each parameter is placed in the Matlab workspace as a double variable An integer corresponding to the selected test is placed in the workspace The simulation script is executed after this the simulation data is collected from the Matlab workspace and plotted Because the exact number of parameters or outputs is not fixed loops were used A wait bar see Figure 3 6 is provided to remind the user that the simulation is in process To ensure that the wait bar remains on top of all other Matlab windows the window should be modal May07 18 Detailed Design Page 37 of 65 Simulating KA Figure 3 6 Wait bar 3 2 3 1 1 Features Several user friendly features have been added A summary of these features and their functionality appears in Table 3 3 Table 3 3 Product Requirements and Features pm er AP Supplemental Intended Functionality enter a specific value for a parameter Slider Bar Required Provide the user with an easy method to vary parameters Each slider bar shall have a default maximum and minimum value When the slider bar position is changed the value in the edit box shall update If a value is entered into the edit box that is outside of this range the corresponding maximum or minimum should reset to the value in the edit box models from the graphical user interface in the graphical user interface Import Data from Excel and Required The user shall be able to import time
78. to the edit box values entered should include values inside and outside the minimum and maximum range specified in the setting file Place a breakpoint inside code to verify that the minimum and maximum values are properly reset when value outside the minimum and maximum range is entered Expected Result Slider bar should move accordingly to the entered numbers the minimum and maximum values should reset properly when a value outside of the default range is entered Actual Result R2006a Edit boxes act correctly R2006b Edit boxes act correctly Action Required none Completed Yes Test Case Enter invalid value in edit box Purpose To verify the program gives proper warning message and continues to function properly after an invalid non numeric value is entered in the edit box Procedure Enter various non numeric values into the edit box Expected Result Warning message of invalid value should appear and the program should continue to function properly Actual Result R2006a Same as R2006b R2006b Entering invalid value resets the edit box to previous value Action Required This is acceptable none Completed Yes Test Case Change parameter value using sliderbar Purpose To verify sliderbar is implemented properly Procedure Move slider between min max and verify value inside edit box is being updated accordingly May07 18 Approach and Results Page 19 of 65 Expected Result Value inside edit box updates p
79. ut the need to change code to the top level program Additionally the close relationship between the team and the client allowed a continuing process of product improvement throughout the implementation process Several key features not originally discussed were added to the program during the implementation that improved functionality and usability Some of these functions were batch simulation processing data exportation and undo and redo functions 2 1 5 End Product Testing The overall testing process will be broken down into two parts e Model testing e Functional testing Model testing outlines the procedures to validate the correctness and accuracy of the Simulink models Functional testing outlines the procedures to verify that all functional requirements and features have been implemented correctly 2 1 5 1 Model Testing The purpose of model testing is to verify that e The Simulink models provide accurate results e The Simulink models are able to provide results over the entire normal range 2 1 5 1 1 Scope and Approach Model testing will include all testing activities necessary to verify that the models produce accurate and reliable results Each model included in the library by March 21 2007 was tested individually 2 1 5 1 1 1 Models Included in Model Testing The following models were tested May07 18 Approach and Results Page 8 of 65 e Unexcited genrou model e Genrou model with exst4b exciter e Genrou with
80. values of these properties are contained in the GUI setting file and with the exception of the default value may change through out the session If a value is entered into the edit box that is not between the minimum and maximum the range resets to include the new value Other important features in the parameters panel are the simulate and batch simulate buttons and the test selection drop down menu The simulate button places the May07 18 Detailed Design Page 36 of 65 parameters in the workspace executes the simulation script collects and plots the simulation results The batch simulate button allows the user to run several simulations at once with one variable being changed each time This allows the user to visualize the effects of one parameter on the simulation results The test selection menu allows for several different events to be incorporated into one model and display The models built thus far have three different tests a step change in reference voltage a load rejection test in which the main breakers are opened and a three phase fault Each time that the simulate button is pressed the entry selected in the test menu is placed into the workspace as an integer this is then collected by the simulation script and a switch statement determines which initial conditions to use and test to run Because in real world testing situations different tests may require different initial conditions the graphical user interface
81. yed the correct default value Action Required none Completed yes 2 1 5 1 2 4 Verification of Maximum Edit Box Values Test Case Verify maximum edit box values Purpose To verify that the default maximum edit box values match the data sent by General Electric Procedure Open each model from the opening window and use the slider bar to set all values to the maximum Compare the edit box values to the data sent by General Electric Expected Result All maximum edit box values for all models should match the data sent by General Electric Actual Result R2006a All included models displayed the correct default value R2006b All included models displayed the correct default value Action Required none Completed yes 2 1 5 1 2 5 Verification that Changing Parameters Produces Different Simulation Results Test Case Verify that changing the edit box produces different simulation results Purpose To verify that changing the edit box produces different simulation results Procedure Open each model from the opening window and press simulate One by one use the slider bar to set a parameter values to a value that is very different from its default value Press the simulate button and use the previous simulation trace to verify that at least one of the curves changed May07 18 Approach and Results Page 11 of 65 Expected Result All slider bars corresponding to parameters other than those belonging to saturation points will pro

Machine Model Parameter Determination Final Report

Contents

Download Pdf Manuals

Related Search

Related Contents