Group Project Description Design And Implementation Of MPC .

2y ago
16 Views
2 Downloads
928.47 KB
7 Pages
Last View : 10d ago
Last Download : 3m ago
Upload by : Aiyana Dorn
Transcription

Group Project, Model Predictive Control (IIA4117)Roshan SharmaGroup Project DescriptionDesign and implementation of MPC for controlling two degrees of freedom (2-DOF)helicopter unit1.Introduction/InformationThe group project for the course IIA4117 (Model Predictive Control) is an obligatory task that all the studentsenrolled in this course should fulfill. It counts 40% of the course grading. This is a group project and eachstudent should be a part of a group. The students have the responsibility to form the groups.A technical report should be delivered by each group (not by each individual student). Each group should alsodemonstrate/implement their controller to the real helicopter units at USN. The deadline for submitting thereport and the date for the demonstration is given at the end of the course homepage(https://home.usn.no/roshans/mpc) (scroll to the bottom).2.Process descriptionThe 2-DOF helicopter unit developed at USN is a dynamic system with multiple inputs and multiple outputs(MIMO). Figure 1 shows the schematic of such a unit with the side view and the top view. It consists of twopropellers (pitch and yaw) driven by motors.Figure 1: 2-DOF helicopter unit at USN (side view and top view)The unit has two inputs: (a) voltage to the front or pitch motor/propeller system, and (b) voltage to the backor yaw motor/propeller system. When voltage is applied to the pitch motor, the pitch propeller rotates and itgenerates thrust, and the helicopter lifts up. Thus voltage to the pitch motor/propeller control the elevation(or pitch) of the helicopter nose about the pitch axis. When voltage is applied to the yaw motor, the yawpropeller rotates and it generates torque in anti-clockwise direction, and the helicopter rotates about the yawaxis. The angle between the pitch axis and the helicopter body axis is called the pitch angle. The angle betweenthe yaw axis and the helicopter body axis is called the yaw angle. The pitch and the yaw angles are measured

Group Project, Model Predictive Control (IIA4117)Roshan Sharmaby using the angle sensors as shown in Figure 1. Thus, these are the two outputs of the system which aremeasureable.Let us define the following:Inputs:π‘‰π‘šπ‘ voltage applied to the picth motorπ‘‰π‘šπ‘¦ voltage applied to the yaw motorOutputs:πœƒ pitch angleπœ“ yaw angleThe process is a cross-coupled MIMO system. When sufficient voltage is applied to the front motor, thehelicopter not only pitches up but it also starts to rotate at the same time i.e. the input π‘‰π‘šπ‘ affects both outputsπœƒ and πœ“. Similarly, when sufficient voltage is applied to the back motor, the helicopter rotates in the anticlockwise direct and at the same time, it also changes its pitch a little i.e. the input π‘‰π‘šπ‘¦ affects both outputsπœƒ and πœ“. The effect of π‘‰π‘šπ‘ on πœ“ is very strong denoted by strong cross-coupling in Figure 1, while the effect ofπ‘‰π‘šπ‘¦ on πœƒ is weak denoted by weak cross-coupling.The system can be described with four states:πœƒ pitch angleπœ“ yaw angleπœ”πœƒ pitch angular velocityπœ”πœ“ yaw angular velocityThe block diagram of the system showing the inputs, outputs and the states is shown in Figure 2. For detaileddescription of the helicopter see the video athttps://web01.usn.no/ .mp4.Figure 2: Block diagram of the system showing inputs, outputs and states

Group Project, Model Predictive Control (IIA4117)Roshan Sharma3. Mathematical model of the processThe dynamics of the 2-DOF helicopter unit is modelled using Newton’s laws of motion and Euler-Lagrangeequations. The model is described by a set of four ordinary differential equations (ODEs).π‘‘πœƒ πœ”πœƒπ‘‘π‘‘π‘‘πœ“ πœ”πœ“π‘‘π‘‘(1)𝐾𝑝𝑦 π‘‰π‘šπ‘¦πΎπ‘π‘ π‘‰π‘šπ‘π‘‘πœ”πœƒ 22π‘‘π‘‘π½π‘’π‘ž,𝑝 π‘šβ„Žπ‘’π‘™π‘– π‘™π‘π‘š π½π‘’π‘ž,𝑝 π‘šβ„Žπ‘’π‘™π‘– π‘™π‘π‘š22𝐡𝑝 πœ”πœƒ π‘šβ„Žπ‘’π‘™π‘– πœ”πœ“sin(πœƒ) π‘™π‘π‘šcos(πœƒ) π‘šβ„Žπ‘’π‘™π‘– 𝑔 cos(πœƒ) π‘™π‘π‘š 2π½π‘’π‘ž,𝑝 π‘šβ„Žπ‘’π‘™π‘– π‘™π‘π‘š2π‘‘πœ”πœ“2π‘šβ„Žπ‘’π‘™π‘– πœ”πœ“ sin(πœƒ) π‘™π‘π‘šcos(πœƒ) π‘€πœƒπΎπ‘¦π‘ π‘‰π‘šπ‘πΎπ‘¦π‘¦ π‘‰π‘šπ‘¦ 222222π‘‘π‘‘π½π‘’π‘ž,𝑦 π‘šβ„Žπ‘’π‘™π‘– cos (πœƒ) π‘™π‘π‘š π½π‘’π‘ž,𝑦 π‘šβ„Žπ‘’π‘™π‘– cos (πœƒ) π‘™π‘π‘šπ½π‘’π‘ž,𝑦 π‘šβ„Žπ‘’π‘™π‘– cos (πœƒ) π‘™π‘π‘šπ΅π‘¦ πœ”πœ“ 2π½π‘’π‘ž,𝑦 π‘šβ„Žπ‘’π‘™π‘– cos2 (πœƒ) π‘™π‘π‘š(3)(2)(4)There are altogether 11 parameters of the system. π‘™π‘π‘š is a parameter which denotes the distance of the pivotpoint from the center of mass as shown in Figure 3. The helicopter rotates freely (up/down and clock/anti clockwise) about the pivot point. Center of mass is the point at which the gravitational force act on the system.Figure 3: Illustration of center of mass and pivot pointAll the helicopter units at USN are fastened and screwed in a way that π‘™π‘π‘š 1.5 [π‘π‘š]. For a fixed value of π‘™π‘π‘šand π‘šβ„Žπ‘’π‘™π‘– , the moment of inertia can be calculated and have been listen in Table 1.Table 1: Parameters of the system: part 1

Group Project, Model Predictive Control (IIA4117)Roshan nce between the pivot point and the center of mass ofhelicopterTotal moving mass of the helicopterMoment of inertia about the pitch axisMoment of inertia about the yaw axisAcceleration due to gravity on planet [kg-m2][m-s-2]The remaining 6 parameters of the system namely, 𝐾𝑝𝑝 , 𝐾𝑦𝑦 , 𝐾𝑝𝑦 , 𝐾𝑦𝑝 , 𝐡𝑝 and 𝐡𝑦 are slightly different foreach unit of the helicopter at USN. For one of the unit, Table 2 lists the values of these parameters. You canuse the parameters listed in Table 2 for simulations. But when you apply it to a real helicopter unit (the onethat is assigned to you), make sure that you fine tune the parameters if necessary.Table 2. Parameters of the system: 𝐾𝑦𝑝𝐡𝑝𝐡𝑦DescriptionTorque constant on pitch axis from pitch motor/propellerTorque constant on yaw axis from yaw motor/propellerTorque constant on pitch axis from yaw motor/propellerTorque constant on yaw axis from pitch motor/propellerDamping friction factor about pitch axisDamping friction factor about yaw ][Nm/V][Nm/V][N/V][N/V]4. Miscellaneous information when using real helicopter unit at USNThe rectangular white box attached to the base plate in Figure 1 is used for data acquisition. Inside this box, adata acquisition device from National Instrument, NI-USB-6001 is present. This is used for reading the pitchand yaw angles of the helicopter, and for sending the control voltages to the pitch and yaw motors. There arealso other components like the power supply unit and Arduino inside the white box. Reading and writing signalsis via USB Plug n Play feature. There is a USB cable connected to the data acquisition box. This USB cable canbe directly connected to a computer for acquiring and sending data from/to the helicopter.4.1Reading pitch and yaw angles of the helicopterThe analogue input channel of the data acquisition can be used to read the pitch and yaw angles of thehelicopter. The angles are read as voltage signals and NOT directly as angles in radians. Thus, these voltagesignals have to be converted to actual angles in radians using linear relationship.Channel ai1 :Pitch angleRead as reference single ended voltage in the range [0 5] voltsIt corresponds to angles in the range [-0.7854Channel ai0 :0.7854] radians (πœƒ)Yaw angleRead as reference single ended voltage in the range [0 5] volts

Group Project, Model Predictive Control (IIA4117)Roshan SharmaIt corresponds to angles in the range [0 πœ‹] radians (πœ“)4.2Applying voltages to the pitch and yaw motorsThe analog output channel of the data acquisition can be used to apply the voltages to the pitch and yawmotors.Channel ao1 : voltage to the pitch motor or front motor, (π‘‰π‘šπ‘ )Sent as reference single ended voltageCareful: The range of π‘‰π‘šπ‘ should be [0 3] volts. Applying voltages outside of thisrange might damage the pitch motor.Channel ao0 : voltage to the yaw motor or tail motor, (π‘‰π‘šπ‘¦ )Sent as reference single ended voltageCareful: The range of π‘‰π‘šπ‘¦ should be [0 3] volts. Applying voltages outside of thisrange might damage the yaw motor.However, when you run a helicopter unit, it is not necessary to put a voltage more than 2 V to the frontmotor (Vmp). With Vmp 2 V, the helicopter unit would have already lifted up. Similarly, it is not necessaryto use a voltage greater than 2 V for the back motor during the operation.4.3Correct way of starting up the helicopterBefore you turn ON the power switch button located at the side of the data acquisition box (the whiterectangular box), the helicopter should be placed at the HOME position. The home position is when πœƒ 450 π‘œπ‘Ÿ 0.7854 [π‘Ÿπ‘Žπ‘‘], and πœ“ 00 π‘œπ‘Ÿ 0 [π‘Ÿπ‘Žπ‘‘]. This is shown in Figure 4.Figure 4: The home position of the helicopter unitHome position: With white data acquisition box facing away from you, rotate the helicopter anti-clockwise toits farthest end and let it freely rest.After the helicopter is at the home position, turn on the power switch button. Wait for beeping sound fromboth the pitch motor and the tail motor (you will hear two beeping sounds, sometimes together and othertimes with a little delay between the two sounds). After the beeping sound, the system is ready to be used.Note: Only after the beeping sound is finished, you should β€œRun” your simulation program in Simulink.

Group Project, Model Predictive Control (IIA4117)Roshan Sharma4.4Use it with CAREThe helicopter units at USN are fragile and not very strong mechanically. Please do use the units with care. Donot put mechanical stress. Avoid operating them such that they bang and reflect back. You can use your handto catch it and to stop the mechanical stresses. Not much attention have been given to its structural strengthduring its design.5. TASKSThe task will be divided into two major parts:(i)(ii)5.1Development of simulatorImplementation to a real helicopter unitSimulator developmentFor developing simulator(s), you can work on your computer(laptops) with Simulink/MATLAB. You donot need access to the real process. The following tasks should be completed:(i)Describe the 2 DOF helicopter process. E.g. How does the process operate? What are theinputs, outputs and states of the system? You can re-use the figures/pictures that is availableto you or you can take/make your own figures.(ii)List all the equations involved with the mathematical model of the process. Describe a littleabout the mathematical model. E.g. What kind, type? What is its order? linear or nonlinear etc.(iii)Using Simulink, perform the openloop simulation of the nonlinear model of the process.Manually excite your system with different values of the control inputs and see the responseof the system. Plot the necessary variables of interests (e.g. angles, velocities, control inputsetc.). Discuss your simulation results.(iv)Choose a suitable operating point and linearize the model of the process about this operatingpoint. Show all the necessary calculations involved during the linearization of the mathematicalmodel.To linearize your model, first choose a suitable operating point for the four states (for e.g.( (v)(vi)10πœ‹ πœ‹, , 0,0) for pitch angle, yaw angle, pitch velocity and yaw velocity respectively). To find180 2the corresponding operating point for the control inputs, solve the nonlinear process model atthe steady state for π‘‰π‘šπ‘ and π‘‰π‘šπ‘¦ using the operating point for the states, i.e. set the left handside of equations (1) to (4) to zero solve these equations (by using the operating point of thestates) for calculating operating point for control input voltages.Using Simulink, simulate the linear model alongside the nonlinear model. Compare thebehavior of the process from these two models. Comment and discuss about your simulationresults.Assume that you can measure only the two angles (the pitch angle (πœƒ) and the yaw angle (πœ“)).Design a suitable state estimator in Simulink for estimating the remaining two states (the pitchvelocity (πœ”πœƒ ) and the yaw velocity (πœ”πœ“ )). Here you are not limited to using steady state Kalmanfilter but you can use any other state estimator that you know if you want. Change the control

Group Project, Model Predictive Control (IIA4117)Roshan Sharma(vii)(viii)inputs manually and observe the behavior of your state estimator. Discuss and comment indetail about your simulation results.Design an output feedback linear MPC for controlling the pitch angle (πœƒ) and the yaw angle (πœ“)to their respective setpoints. At this point, do not worry about integral action, however, pleasedo consider the following:a. Your controller should not supply more than 3[V] as the control inputs. They cannot be lessthan 0 [V] either. Using signal limiter block in Simulink is NOT ENOUGH. You shouldconsider this while formulating your optimal control problem.b. You should show the MPC problem formulation very clearly. Show the objective function,constraints, bounds etc. Show your choice of the unknown vectors to optimize.c. You should show every detailed step of how you calculated/formulated all the necessarymatrices and vectors. How did you handle the bounds (if any?)d. How did you implement the receeding horizon strategy in Simulink? Show clearly.e. The linear MPC should be applied to the helicopter process. For the simulator, considerthe nonlinear model of the helicopter as your process. Please DO NOT use the linear modelas your process. You can of course make use of things that you already did in task (iii) forthe nonlinear model of the process.f. Provide step changes to the setpoint values and observe the response of your controller.Change your setpoint as ramp (for e.g. using slider or knob in Simulink) and observe theresponse of your controller.g. Plot all the necessary variables (outputs, estimated states, inputs, references etc.)h. Discuss in details all your simulation results.i. Among many other things, did you observe any steady-state offsets with your linear MPC?Design an output feedback linear MPC with integral action. Consider the points (vii) a, e, f andg while designing the controller. How can you ensure integral action with your linear MPC? Ifyou know many ways to achieve integral action, just describe the one that you have used inyour project. Discuss in detail your simulation results.5.2 Implementation on a real helicopter unit.After you think that your simulators are working as expected, you can then implement your linear MPCon a real helicopter unit. The campus students have a golden opportunity to use the units at USN rightfrom the beginning of the project work. For the online students, implementation/testing of youralgorithms should be performed on the designated dates (see the course home page (scroll to thebottom) for the actual dates). Please fulfill the following:(i)Implement the state estimator on the real helicopter unit for estimating pitch velocity (πœ”πœƒ )and the yaw velocity (πœ”πœ“ ). Plot all the necessary variables (e.g. estimated states, outputs,inputs etc.). Change the control inputs manually and observe the behavior of your stateestimator. Discuss and comment in detail about your results.(ii)Implement a linear MPC with integral action to the real helicopter unit. Provide step changesto the setpoint values and observe the response of your controller. Change your setpoint asramp (for e.g. using slider or knob in Simulink) and observe the response of your controller.Plot all the necessary variables (outputs, estimated states, inputs, references etc.). Discuss indetails all your results.Good luck!

data acquisition device from National Instrument, NI-USB-6001 is present. This is used for reading the pitch and yaw angles of the helicopter, and for sending the control voltages to the pitch and yaw motors. There are also other components like the power supply unit and

Related Documents:

Source: 2016 Miami-Dade County Infant Mortality Analysis Highest Neighborhood Rates and Percentages. 31 21.6 23.7 26.26 37.84 41.84 42.28 43.74 50.7 54.06 58.9 64.56 68.34 77.04 0 20 40 60 80 100 Group F Group D Group N Group G Group B Group H Group C Group M Group J Group A Group E Group K Group I Percent p

U8 Whitecaps Jan Levius Monday (5:00 Group A / 6:00 Group B) Field 11A Thursday (5:00 Group A / 6:00 Group B) Field 10A U8 Sounders Greg George Tuesday (5:00 Group A / 6:00 Group B) Field 10A Wednesday (5:00 Group A / 6:00 Group B) Field 9B U8 Red Stars Ty Hesser Monday (5:00 Group A / 6:00 Group B) Field 10B Thursday (5:00 Group A / 6:00 Group .

TAM 251 Group Project: design of a swing set. Fall 2015 1. Project description: This semester, your team will design a playground set to be installed in my own backyard next Spring. My daughter was so excited when I told her that I have a team of 250 engineering students coming up with an amazing new design! Your design constraints are:

Title of Innovative Project: Name of Project Lead/Team: School Phone Number Years in District Project Description . Project Description – Describe your project in 200 words or less. Include the problem statement, purpose for the project, project goal, intended outcome(s) and a simple overview of proposed plan of action to address the problem.

Output: Project management plan and project documents . 15 . Actions Involved in the Planning Process Group The project scope statement includes: - Description of the project scope - Project deliverables - Project constraints - Project assumptions Work Breakdown Structure (WBS):

Project Design Guidance provides a comprehensive approach to project design that is relevant to all ChildFund supported projects. The processes and steps in this Guidance represent minimum standards for project design. The Project Design Guidance will help ensure consistent quality of project design across all ChildFund supported projects.

2 CHANNEL Group 11 Group 12 Group 13 Group 14 Group 15 Group 16 Group 17 Group 18 Group 19 Group 20 1 472.225 470.300 470.500 478.200 486.200 494.200 470.125 470.575 470.525 470.350 2 472.975 472.225 471.400 478.775 486.775 494.775 472.000 472.100 471.575 471.125 3 476.700 477.100 471.925 480.100 488.100 496.100

mitsubishi electric wind baffle - wb-pa3 mr. slim p-series pumy-60 submittal wb-pa3 description description c description d description e description f description g description proprietary and confidential the information contained in this drawing is the sole property