6 - Not Found
Contents
1. able to receive data from them The entire control algorithm from sensor input to thruster allocation was implemented The sensor data was properly passed to the control algorithm and used effectively Control code for some of the actuators was written A rudimentary graphic user interface GUI was created that displayed all the sensor data and some of the control data The GUI was implemented on a Windows laptop provided by Titanic Positioning Goals to Complete The project will be continued during the summer by a couple members of the team under the employment of Think Sensor Research The option will be available for other members to join during the fall semester as either a coop or full time employment The actuators will be connected to the control algorithm and will be controlled by the output data The Graphic User Interface will be improved to show data more clearly and to allow inputs from the user A joystick will be added for user input The project will be upgraded from a proof of concept to a production version The control code will be optimized and streamlined The display will be presented on a marine display package The current embedded computer will be replaced with a marine rated embedded computer with built in serial ports The GPS will be upgraded to a marine grade GPS A north seeking gyroscope will be used to acquire more accurate heading data than the MRU 9 Appendix A Meeting Minutes Titanic Position2. implement the wind feedforward and thruster allocation part of the control system For the thruster allocation part I had to learn how to linearize a non linear matrix and how to invert the linearized matrix I also developed the skills necessary to implement the mathematical algorithm in C I also learned how to multithread input sensor data This is important because the input sensors all send data in at different rates The MRU is the fastest and the wind sensor is the slowest If the code is single threaded a latency is built up on the MRU data since it has to wait for the other sensors to finish updating To solve this problem I learned to create a thread for each individual sensor to allow multiple operations to be done simultaneously Then when the thread read the data it locks off the processing function so that there are not conflicting processed data sets 7 2 Bardia In this project my roll was mostly the serial IO interfacing through Linux kernel functions I learned a lot about asynchronous serial communication and RS232 handshaking protocols I found this experience very valuable because if one technical skill that I can say I learned from 440 was interfacing hardware and software I learnt a lot about how to parse serial inputs and how to deal with data synchronization Second skill I can take away from capstone is testing and debugging I learnt about significance of these skills and how they play a major role in the design
3. tasks For example when reading the Fossen textbook I spent time trying to fully understand each part of the controller algorithm rather than just looking at the big picture When doing project management work in the future I would read enough to assign specific tasks and let each person find the technical details for their specific task This would speed up the process and allow people to start working on there specific tasks earlier Then it would leave room for going back and learning the technical details later and being available to help team members understand the technical information Learning the high level design first and not getting stuck on learning the low level details is important because it allows team members to contribute more easily From a technical perspective I learned how to program a control system in the C language I picked up many important skills such as finding and linking libraries that perform certain functions and simplify the code for example using the Eigen matrix library allowed for easier matrix math operations Using established libraries makes certain operations more efficient and reliable since it has been widely established Itis also important to make sure that all group members use the same library for the same operations because this makes integrating the code more efficient In this project I learned how to implement and design a control system using matrix math operations More specifically I learned to
4. the data that is required for the estimation requires data and tuning from the boat which we did not end up having access to in time I plan on continuing working on the state estimator in the summer and taking more time to understand the theory well 7 5 Yalda In this project I was mostly involved in data parsing and programming the controller It took me a while to understand how to interpret data from the sensors and how to communicate with them through serial ports It was really exciting and challenging at the same time because I never actually got a chance to do a similar work during school Therefore I had to spend a good portion of my time doing research about the design procedure Even though I have not taken the embedded system course yet working on this project gave me a good opportunity to understand a lot of concepts about embedded systems in advance For instance I was able to understand what an embedded system does and how I can communicate with several components through an embedded system and interpret the data I obtain from it I also worked on making a digital filter for the control system This also gave mea good opportunity to refer to what I had learnt from the related course In that course however we made a filter on a bred board so I had never experienced coding a digital filter in C Therefore I had to do a lot of research on how the digital filters work where to start and how to interpret the output from a
5. ENSC 440 Post Mortem Dynamic Positioning Control System Group 5 Bardia Bogharti bboghrat sfu ca Bengt Haunerland bkh4 sfu ca Lucie Hiltner Ihiliner sfu ca Yalda Majdi ymajdi sfu ca Carl Wahlstrom cwahlstr sfu ca uM ma Iu TITANIC POSITIONING Table of Contents L aSEOPEISUEO SS eo iidem idisse laci d eni Free oko iren Mmi D rese RE RE Ert nae ERE en ERR rU RENE 2 IL Lastof Tables uec AARAL 2 1 riseiirusture ese ABNER cba ct cdc maaana ada eb ataen nies cd pede aiik annesi aia 3 2 Main HINCUONS anie eoe Midi naanakan Ana nahin Gan Na 3 3 COSEANALYSIS ia L 7 4 Schedule comparison aic c daten na SCIO RAO Re ERA FLA DA RR Co EL A 9 Ba Challenges is teed cR 10 br Group DYNAMIC chinese ta cin casas ht ann co tice da wasabi cides aunt e A 11 Ze ROMO CH OM vaserinisseicessnscessnivcadeinienscecedeseccaretewenseieniiensuneessadebasicaveeenntinaneiideieenenintenniennineessiiiandninieannt 11 T Belg swathes ut sacar seca sis tlc KANE A tu MCA Dd C aa 12 PAMAN CGE AA AA cia CLR RIT cia us i AA 13 SEE anakan NAAN OK EN RD TE 13 4 RITU mI c 15 AS Inr E 16 8 Odit Der 17 9 Appendix A Meeting Minutes iieececiienneasiicicarac idus cra dia Xa N kei DX XT NR FUE C End E S 18 I List of Figures Figure 2 1 Flowchart showing data streams for DPS 0 ss 5 Figure 2 1 Flowchart of the DPS controller The colour coding matchesFigure 2 1 M M
6. at it is better to just continue working on an unfinished goal rather than become angry at someone who has not contributed as much as originally desired I came to truly appreciate the need to fail faster We had some issues waiting for hardware to arrive We should have filled that time more with writing internal components ofthe controller that were not affected by the sensors or actuators Had we done this we would have found internal bugs with our algorithms and math for the controller sooner and there would not have been as much of a crunch to complete the project once we were integrating the hardware Ilearned that a journal can be a useful device It can help me keep track of the timeline of a project when it all started to blur together It was also a useful source of scrap paper and spit balling ideas I realized that the journal does not have to bea pristine document clinically documenting my progress and sterile The first hurdle we faced were the various documents During these I learned what kind of audience to write for and what style to write in or sometimes the opposite when constructing a technical document The first major technical problem we faced was setting up the IDE environments before coding Finding appropriate C libraries and then configuring the environment and code so we could utilize them took a lot of time It was something none of us had much experience with and something that engineers are never taught i
7. avel showed us the batteries Marine Deep Cycle 12V we can use for mobile testing at SFU Located in URL Bardia Bengt and Carl worked on the Functional Spec document in Lab 1 a little bit Internal deadline for FS is Friday Feb 14 Pavel would like a copy of FS so he can send it to his contacts Suggest including Pavel during weekend review edit of FS Titanic Positioning MINUTES February 18 2014 4 00 4 30 Underwater Research Lab Present Bengt Carl Lucie Absent Yalda Pavel Think Sensor Reseach Bardia Location Underwater Research Lab Purpose of Meeting To discuss next steps of software developing Minutes Bengt called the meeting at 4 A discuss next steps for software development Discussion Action 1 Set up linux environment Find and install drivers for MRU GPS wind sensor angle sensor and autopilot interface 3 Plug in sensors and confirm communications 4 Revisit and create a plan for group coding B discuss missing hardware Discussion o Wind sensor o Angle sensor either potentiometer or digital encoder o Motor throttle control Action Once equipment parts have been found email Pavel the list and he will order the parts C establish next steps Discussion research operating systems that we want to use for the embedded computer Options linux embedded linux windows embedded windows QNX Needs to be compatible with processor of embedded computer Needs to have device drivers readily available
8. d sensor information gathered from wind and water current data The purpose of the project is to implement a DPS on a 10 meter long boat The boat will maintain position and heading against external forces such as the current in the Fraser River The system will be expandable and with the addition of further software will be able to implement full dynamic positioning The proposed requirements for the DPS are defined in this Design Specification 2 Main functions 2 1 Sensors and Inputs This section lists the sensors that are used in the project The following sensors are required to implement a DPS 0 GPS MRU joystick and emergency stop button The GPS MRU and anemometer are used for collecting data The joystick is used for controlling the desired location and heading The emergency stop is used for safety to manually stop the DPS Table 2 1 below shows the sensors and inputs that will be used in the proof of concept design Sensors and Inputs Anemometer 6 Joystick Emergency Stop button 7 Product Number MD 762 1009 07A SR 0404 14391 0001 TSR 100 Garmin GWS 10 X03 57540 Game Switch SPDT Square Red Lens On On Illuminated Table 2 1 Products used for proof of concept Specifications Serial RS 232 Average GSP resolution is 1 m Anodized aluminium 7 36 Vdc power 1cm heave resolution 596 heave accuracy 0 5 degree orientation accuracy serial RS 232 Wind speed and w
9. digital filter This project also gave me a good opportunity to develop my coding skills further I gained experience on how to find useful libraries and functions online and link them to my code to simplify the coding procedure It also gave me an opportunity to refresh my memory on C coding techniques and syntaxes In terms of personal skills I learnt how to manage my time properly How to deal with stressful situations where we were very close to deadlines and still had not finished the required task I learnt that the easiest way to avoid stressful situations was to plan well in advance Also I learnt that it is always good to have a plan B in case your first plan does not work as expected I was also able to improve my teamwork skills I received a lot of help from the other team members when I was stuck on a task and I also tried to help others as much as I could Working on this project made me believe that you do not really have to know every details of a project in order to improve it There were times that I had to spend so much time doing research and working on a piece of code in the end however I noticed that I was not on the right track and that had to start it from the beginning I learnt that even in these situations I could gain a lot of knowledge and experience that could be useful later 8 Conclusion Goals Met The groundwork for a DPS 0 was firmly established The sensors were properly implemented and the system was
10. docs o Skype facetime google hangouts o Dropbox Action Blake will email group the email he uses for dropbox and google docs D Next Meeting Date The next meeting was arranged for January 14 2014 at 12 30 2 30 in the underwater research lab Meeting was adjourned early at 11 00 Titanic Positioning MINUTES January 14 2014 12 30 1 30 Underwater Research Lab Present Bengt Carl Lucie Yalda Pavel Think Sensor Reseach Absent Bardia Purpose of Meeting To discuss the purpose of the project and establish logistics on cost Minutes Pavel called the meeting at 12 30 A Discuss scope of project To get a manual control DP system running on a boat in Barnett Marina Discussion Consider options for how to implement the system Action o Option I implement using PLC needs floating point operation o Option 2 implement using a embedded computer o Enter options into proposal and decide later which is better B discuss equipment Discussion discuss list of parts needed Marine GPS Gyroscope Motion reference unit Joystick for controller serial port O O O Action Once equipment parts have been found email Pavel the list and he will order the parts C establish next steps Discussion the following actions to be completed after proposal is done o Get on the boat to see what we have to work with o Possibly rewire the boat if needed o Get list of equipment put together D Next Meeting Date The next m
11. e Since we were programming the DPS for Think Sensor Research I learned about the design process and aspects of the design that were required to bring a product to market Designing a practical system that can be implemented in industry and having the support to do so is the reason why this project was so appealing to me in the first place One of the difficulties we faced was the time it took to order and receive the products I learned that relying on others for critical parts of the project is necessary and will often cause the project to progress at a slower rate than I would like I learned that it s important to plan ahead for these wait times and in the future to plan out what can be worked on when parts this could extend to parts of the code not just hardware are not available Another thing I learned that I had not done before was designing with standards in mind in this case from the American Bureau of Shipping This project gave me a realistic look at what it will look like when I am working in industry and will help me with project planning and time management in the future The first thing I had to do was research navigation principles This took a long time and lots of reading but I learned about a whole area of engineering that I had not considered before Designing the PID controller was challenging I spent a long time learning about control systems control theory and designing and tuning PID controllers Despite the long ti
12. eeting was arranged for January 20 2014 to fine tune proposal Meeting was adjourned early at 1 30 Titanic Positioning MINUTES February 3 5 2014 12 30 1 30 Underwater Research Lab Present Bengt Carl Lucie Yalda Pavel Think Sensor Reseach Bardia Absent none Location Underwater Research Lab Purpose of Meeting To get information on the Motion Reference Unit and discuss next steps Minutes Pavel called the meeting at 12 30 A Discuss MRU GPS and other hardware MRU is made by Think Sensor Research and is supplied by Pavel Discussion Discuss hardware Action MRU was brought by Pavel and is stored in the Underwater Research Lab MRU user manual provide with specs of MRU Discussed joystick needs to be 3 axis and have USB connection Consider using CBC connectors IP67 rated for water spray Should use RS 232 serial connectors for the proof of concept Wind speed needs to be serial since USB does not have long enough cables QO O O O O O B discuss equipment needed for interfacing with the boat Discussion o Hydraulic steering auto control will be supplied by Pavel since the hydraulic steering control is out of scope o Throttle control still needs to be addressed Action Once equipment parts have been found email Pavel the list and he will order the parts C establish next steps Discussion research operating systems that we want to use for the embedded computer Options linux embedded linux windows embedded w
13. f the DPS controller The colour coding matches Figure 2 1 The proof of concept design uses a development board with a separate USB to serial adapter The USB to serial adapter will be enclosed in a metal container with the wires temporarily secured The model doesn t use a joystick or a state estimator as originally planned The joystick was replaced with keyboard control since the production model will have an integrated display computer and joystick therefore making it impractical to deal with separate driver installation for the joystick on the current model The keyboard can model the joystick very well The state estimator very difficult to model and is not a requirement for the proof of concept model The code for the state estimator could take years to finish and therefore will be added during summer work terms 3 Cost Analysis Cost Estimate Final Cost TSR 100 MRU loaned from TSR sl Embedded Marine 1500 1500 Computer Clipper Wind Sensor 1129 Captain s time Motor Controller for steering Hydraulic 220 220 motor Linear actuators to attach to throttles Table 3 1 below shows the financial breakdown of the project Some of the components have been loaned from Think Sensor Research TSR TSR financed all other components The costs for cables connectors and use of the boat are approximate Cost Estimate Final Cost TSR 100 MRU loaned from TSR Nu Embedded Marine 1500 1500 Computer Clipper Wi
14. grate and Debug 04 1 2014 04 15 2014 Implement on boat 04 1 2014 TBA Table4 1 Comparison of initial milestones and current progress The MRU and GPS were available by the planned date The embedded computer and Anemometer were ordered later 5 Challenges The embedded computer took longer then expected to arrive We were able to use a loner development board this allowed us to could continue to work on the project instead of waiting for the computer to arrive The display provided some technical challenges since we were initially planning to do the display on the embedded computer using the Linux operating system However installing graphic interface programs on the Linux computer became very difficult and time consuming Linux also doesn t support the C graphic interface program which we were more familiar with The solution to this problem became to use a windows laptop for the display and collected data from the embedded computer through an Ethernet cable We chose to do this since the production version would have a windows computer for the user interface and therefore it was not practical to do the display in a Linux environment This allowed us to use C and to circumvent the difficult graphic interface installations 6 Group Dynamic The group was organized in a democratic fashion where everyone opinion was listened to and decisions were made as a group The administrator would identify how the tasks would be spil
15. ind angle measurements Wind speed angle filtering included Microsoft Sidewinder USB input linux compatible 5A 125VAC Plastic pushbutton assembly Justification Re used marine grade Produced by Think Sensor research has a compass and measures heavy pitch roll and yaw has built in filter so that measurements can be take mean or average wind speed Re used has z axis rotation Light gives feedback and is red colour is obvious for stop Figure 2 1 below shows the data streams in the final proof of concept system All the inputs are sent to the embedded computer along serial lines RS 232 or USB ports The controller is contained on the embedded computer and all calculations will be performed on the embedded computer using C C functions The controller interfaces with the monitor and the motor controller unit which is used to control the direction and throttle of all the motors on the boat Bow Thruster Throttle Motor Stern Thruster Controller Throttle Angle Sensor Emergency Stern Thruster Stop Angle Figure 2 1 Flowchart showing data streams for DPS 0 2 2 Controller The goal of the PID controller is to process the signals sent in from the sensors and output a fast and stable response The proportional P component reduces rise time of the response so the DPS reaches its desired position faster The integrator I component reduces steady state error caused by slowly varying ocean cur
16. indows QNX Needs to be compatible with processor of embedded computer Needs to have device drivers readily available not part of scope to write drivers Look for embedded libraries for GPS etc O O O O D Next Meeting Date The next meeting was arranged for February 7 2014 to discuss research for OS and settle on a embedded controller Meeting was adjourned at 1 30 February 11 2014 Minutes Pavel provided us a computer with Ubuntu installed on it and a USB serial splitter We can begin testing sensors and writing code with this machine We need to provide a monitor keyboard and mouse The computer is being stored in Bardia s locker 12 in Lab 1 Ask Bardia for combo Pavel said he would order an embedded computer soon We mentioned that our preferred model does not explicitly support Linux but Pavel was sure that it would Pavel recommended that we put Ubuntu onto a flash drive so we can begin working on the code from any computer 16 Gigabyte drive will apparently hold all the OS IDE and drivers needed Pavel likes the idea of using Linux for the project Perhaps use Windows C for the display computer The display computer will not be required until working on larger ships with longer cables so can be safely ignored for now would more comfortable with C because I ve never used C lm not sure how similar they are lucie An internal deadline for wiring the test boat has been set around March 14 2014 P
17. ing MINUTES January 10 2014 10 30 11 00 McKenzie Cafe Present Bengt Blake Carl Lucie Yalda Absent none Purpose of Meeting To establish a plan for the project proposal and to develop any questions that we have for Pavel Minutes Bengt called the meeting to order at 10 30 A Project Proposal Review Rubric How should we split up the work for the proposal Discussion Establish what can be done before meeting with Pavel and what needs to be discussed with him on Tuesday 12 30 Action Proposal Introduction background Yalda Scope risk benefits Bengt Market competition research rationale Lucie Company details Blake Establish that project planning and cost will be discussed with pavel OO OQ C questions for Tuesdays meeting Discussion o Project planning Develop a plan for the project and exact details on how much of the project will be done in 440 Establish milestones o Cost considerations Develop list of equipment needed to prepare for research on cost o Software What OS will be using linux windows What language c c Will pavel provide licenses for home use or will basic lab computers have the software or lab space available Action Issues will be discussed on Tuesday D establish contact info and programs used for file sharing Discussion e Provide reading material to Blake so that he can get caught up with the overall project e Ensure everyone has access to o Google
18. me I spent learning about control loops the code for the PID controller was relatively simple and not very long Next time I will try to use my time more efficiently I will do this by looking at how people have managed to accomplish a task in the past and try to do something similar to what has already been done as it applies to my project Then I can go back and learn theory that is not so broad and applies only to what I am working on Programming the state estimation part of the controller was the most challenging thing for me technically both because the theory was hard to understand and because it was hard to code The coding involved reading in the coordinates from the GPS and the MRU and using rotation matrices to make the coordinates appropriate for calculation The data from the MRU had to be processed in order to compensate for centripetal acceleration when the boat turns and compensate for pitch roll and heave of the boat The GPS had to be processed in order to compensate for the offset from the GPS to the center of the boat picture the GPS swaying in the air from side to side above the boat in the waves After all the raw data from the GPS and MRU was processed it had to be put into a matrix linearized by differentiation and fed into a Kalman filter for state estimation I didn t complete this part of the code partially due to problems with integration and transforming the filter into c code from matlab and also because most of
19. n any class I am now better at properly storing and linking to libraries In order to eliminate the noise of our output we had to low pass filter the data We attempted to use filter libraries and pre made filters to perform this function for us We realized that all the filters were using the frequency domain while we needed the time domain We began to look into Fourier transforms Eventually we realized that it would be faster to write our own simple discrete time filter which I puzzled out and implemented with no external libraries When we had to create a GUI we originally just used the console window and text files I learned how to manipulate what was displayed and how to use different types of data streams in C Later we decided to implement a GUI on a Windows platform in C I was the only member with any C experience and so I took on the task To create this GUI I had to learn how to use serial port communication protocols as well as creating a dynamic display which I had never done before Overall working with Titanic Positioning was an excellent experience Especially at the end of the project the team came together and we completed many goals There were many moments of celebrations and more than a few groans as we discovered new problems We kept the mood light and remained optimistic though the process and produced an interesting demo I m looking forward to working with these fine engineers in the future 7 4 Luci
20. nd Sensor 1129 Captain s time Motor Controller for steering Hydraulic 220 220 motor Linear actuators to attach to throttles Table 3 1 Financial breakdown of project The total cost for all the equipment as outlined in the table above is 2320 excluding the cost of loaned components Our initial cost estimate was higher because we were expecting to use an ultrasonic wind which is more accurate and more expensive the wind sensor that we ended up using We also did not have to pay for fuel and other boat expenses since we didn t get on the boat 4 Schedule comparison Milestones Planned Due Date Milestone Achieved date Prepare Parts list 01 30 2014 01 31 2014 Inspect boat 01 30 2014 01 24 2014 Order Parts 02 15 2014 03 06 2014 Controller Implementation 03 1 2014 04 1 2014 Input implementation 03 1 2014 04 7 2014 Integrate and Debug 04 1 2014 04 15 2014 Implement on boat 04 1 2014 TBA Table4 1 shows the milestones outlined in the proposal and current progress Unfortunately we were unable to get onto the boat before the demo however installing the system on the SFU test boat is still a plan for the summer Milestones Planned Due Date Milestone Achieved date Prepare Parts list 01 30 2014 01 31 2014 Inspect boat 01 30 2014 01 24 2014 Order Parts 02 15 2014 03 06 2014 Controller Implementation 03 1 2014 04 1 2014 Input implementation 03 1 2014 04 7 2014 Inte
21. not part of scope to write drivers Look for embedded libraries for GPS etc O O O O D Next Meeting Date Friday February 21 Meeting was adjourned at 4 30 Titanic Positioning MINUTES February 18 2014 4 00 4 30 Underwater Research Lab Present Bengt Bardia Lucie Absent Yalda Pavel Think Sensor Reseach Carl Location Underwater Research Lab Purpose of Meeting To discuss progress and integration of software development Minutes Bengt called the meeting at 12 30 A discuss progress of Software Discussion e Wind Feedforward Function completed e Parsing from MRU and GPS completed still need to get data for yaw e PID Loop still need to add matrix support and error function Action e Bengt thruster allocation e Bardia i o of motor controller e Lucie filtering and possibly state estimation also add matrix capacity to PID e Carl Yalda display B Next chance to work on boat and integrate code Discussion e Start wiring boat on Thursday or Friday e Meet on the weekend to integrate code Action confirm that we can get acesss to the boat D Next Meeting Date Meeting was adjourned at 4 00
22. process Also I had to learn quite a bit about compilers to learn various issues that came up in setting up the integrated development environment Having the chance to work on the actuators in our system I encountered a whole side of engineering I have never dealt with before that is the electromechanical devices and their interfacing I learnt about PWM signaling and how different encoders provide digital feedback Aside from the technical skills obtained I also learnt how important it is to stay communicated with your teammates to ensure that task will be integrated easily when design processes is done in parallel among the group mates All in all 440 was a great experience for me during which I was able to gain valuable knowledge about myself engineering and have the chance to make friends and work with some very good people along the way 7 3 Carl One issue with Titanic Positioning initially was team communication Sometimes people were difficult to get a hold of Sometimes team meetings took place without the full team and those absent were not always updated By the end of the project our communication had improved drastically and everyone was working together I learned that the best attitude is a lighthearted one to laugh at setbacks and unforeseen barriers rather than become anxious and stressed I came to properly appreciate that focusing on individual achievement is not a healthy way to approach a project I learned th
23. r TP Error Bookmark not defined ll List of Tables Table 2 1 Products used for proof of concept eene 4 Table 3 1 Financial breakdown of project senes 8 Table 0 1 Comparison of initial milestones and current progress sesser 10 Table 9 1 Shows technical responsibilities Where XX denotes primary role and X denotes secondary role ahead e a hd ARN copas atte ies td 12 1 Introduction Marine vessels are sometimes required to maintain a heading or a position in situations where anchoring is difficult or impractical An automated method of maintaining a heading or position is called Dynamic Positioning Dynamic Positioning is prevalent in the oil industry where large transport ships dock at offshore oil platforms The vessels are required to steadily maintain precise coordinates for hours while cargo is transferred The position of the vessels must be held constant despite exterior forces such as wind waves and tidal pull This must be reliably done to ensure the safety of the workers on the rig and the ship and that no product is spilled in the process 1 A Dynamic Positioning System DPS is a control network that uses the ship s thrusters to adjust and maintain the ship s position according to feedback input from multiple sensors The ship s location is adjusted and maintained by the DPS controller which makes a calculation based on the current location of the ship using a GPS an
24. rents and wave drifts The differentiator D component reduces overshoot so that the system thrusters do not overcompensate Figure 2 2 shows the control system that will be used on the proof of concept The controller is made up of several different components including a PID component a wind feedforward system wave filtering state estimators and thrust allocation The wind feedforward system takes in wind speed and direction from an anemometer and then compensates for the mean wind The wave filtering filters out high frequency wave components from both wind and ocean waves since the boat cannot respond fast enough to compensate for high frequency ocean waves and wind gusts The state estimator is used to estimate unmeasured states such as linear and angular velocities The state estimator is implemented using a Kalman filter The last part of the control algorithm is the thrust allocation which will tell the two stern thrusters which way to turn and what speed to use It will also tell the bow thruster when to turn on and whether to spin clockwise or counter clockwise The output will move the thrusters and provide the new position and heading which is measured by the GPS and MRU The GPS and MRU provide feedback into the system and are compared to the desired input position from the joystick The entire controller system will be contained on the embedded computer 3 E State Da pi it ica Figure 2 2 Flowchart o
25. t up but the group decided together who was best suited for each role There was one instant where 2 members were working on the same section of code and didn t communicate to one another that they were working on the same thing This resulted in 1 person being upset that their hard work was not going to be used However once the issue was identified the group tried to find away to make sure that miscommunication does not happen again The team established a plan that if people start working on something that was not assigned to them during a group meeting then they would email the group to see if anyone had started that part before starting to work on it 7 Reflection Overall the workload distribution was evenly distributed Table 7 1 shows the work distribution for the tasks required for this project Responsibility Bardia Bengt Carl Lucie Yalda Input interfacing XX Data Parsing X XX PID XX Wind XX Feedforward Filtering X XX X XX Thruster XX Allocation Display XX Output XX Interfacing Documentation X XX XX X X Administrative XX X Tasks Table 7 1 Shows technical responsibilities Where XX denotes primary role and X denotes secondary role 7 1 Bengt For this project I took on the project management responsibilities The main thing I learned is that I do not need to know all the details of how things work before I assign certain
Download Pdf Manuals
Related Search