Rotary Inverted Pendulum
Contents
1. J agt mr 0 mLr amp T GL amp mLr mgLa 0 3 12 Referring back to Experiment 1 Position Control we know that the output Torque on the load from the motor is n n KK V KK 9 output R m 3 13 Finally by combining equations 3 11 3 12 amp 3 13 we are left with the following state space representation of the complete system aloo a 0 0 6 0 0 1 0 0 amp 0 bd cG 0 x NaN KK y 6 E E R E m a 0 ae e 0 amp pi KK E E RE Where a J mr E ac b 2 pei alin K KK gt Ba Rn c mL Ry 3 d mgL Page 6 Revision 01 In the typical configuration of the SRVO2 amp the ROTPEN Pendulum inverted pendulum system the above state space representation of the system is 0 0 0 1 olle 0 jjo o 0 aije o f 0 3932 1452 ol 25 54 amp 0 81 78 13 98 0 amp 24 59 3 2 Pre Lab Assignment The purpose of the lab is to design a state feedback controller that will balance the inverted pendulum in the upright position The controller should also maintain the pendulum s stability upright while allowing the user to command varying setpoints of the servo angle The controller specifications for this lab are The controller must maintain system stability The controller must accurately place the servo angle theta at a given command while maintaining the pendulum in the upright position The pendulum angle a must not exceed 6 when the system is excite2. Theta settling time measurements would not give an actual reading and thus not be an accurate indication of controller performance If your controller did not manage to meet the specified requirements you should go back to the simulations and re iterate the design procedure until you have developed a controller that will meet the requirements The previous figures show the WinCon plots used to make the above calculations It should be noted that each sample in the Measurement Table corresponds to one 10 second capture of the plots as seen in Figure 7 amp Figure 8 above You should also capture a plot of Theta as it is required to answer some of the post lab questions Page 13 Revision 01 5 Post Lab Question and Report Upon completion of the lab you should begin by documenting your work into a lab report Included in this report should be the following Vi 5 1 In the pre lab you were asked to theorize as to the range of alpha that you would expect the linear model to be valid Include all calculations and the final estimate In Part Il of the lab you were asked to vary 2 parameters 10 steps for each for a total of 20 entries Make sure to include your Iteration Table in this report With the controller requirements in mind you were asked to determine the optimal parameters to achieve those specifications Include your design steps and all iterations used in determining the final controller After implementing your
3. Q diag 1 1 0 OJ as you can see in the Iteration script At this point you should start varying q amp qz and making note of their effect on the closed loop system characteristics settling time a settling time a range u range By the end of this exercise you should have a table of the form qi q2 Theta 0 Alpha a Alpha Range Control Signal u Settling time s Settling time S Range 1 1 2 55 2 65 3 2 lt a lt 3 2 1V lt u lt 1V 10 1 1 1 1 51 7 2 lt a lt 7 2 3 2V lt u lt 3 2V 1 10 2 55 2 67 2 9 lt a lt 2 9 1V lt u lt 1V Table 1 teration Table Sample Entries Table 1 above shows 3 different entries into the Iteration Table that you should be constructing In total there should be 20 entries in the table vary each parameter in 10 steps while holding the other parameter at 1 Suggested ranges for this exercise are 0 lt qi lt 20 0 lt qe lt 50 Page 10 Revision 01 Now that your Iteration Table is complete look back at the controller requirements of this lab while cross referencing your newly formed table to determine what parameters would result in meeting the specifications Once you determine the best set of parameters that meet the specified requirements you are ready to test your controller on the actual plant Make sure you have properly documented how you obtained your final set of parameters in a table much like the Iteration Table and clea
4. of mass in the y direction Page 3 Revision 01 Pendulum Rotating Arm SRV02 Figure 2 Top View of Rotary inverted pendulum Figure 3 Side View with Pendulum in Motion Figure 2 above depicts the rotary inverted pendulum in motion Take note of the direction the arm is moving Figure 3 depicts the pendulum as a lump mass at half the length of the pendulum The arm is displaced with a given a Notice that the direction of 8 is now in the x direction of this illustration We shall begin the derivation by examining the velocity of the pendulum center of mass Referring back to Figure 3 we notice that there are 2 components for the velocity of the Pendulum lumped mass Lcosa amp Lsin amp 3 1 Pendulum center of mass We also know that the pendulum arm is also moving with the rotating arm at a rate of V r 3 2 arm Using equations 3 1 amp 3 2 and solving for the x amp y velocity components V r0 Lcosa amp 3 3 V Lsna a Equation 3 3 leaves us with the complete velocity of the pendulum We can now proceed to derive the system dynamic equations Page 4 Revision 01 3 1 Deriving The System Dynamic Equations Now that we have obtained the velocities of the pendulum the system dynamic equations can be obtained using the Euler Lagrange formulation We obtain the Potential and Kinetic energies in our system as Potential Energy The only p
5. LQR design q is mainly associated with 0 while q2 is mainly associated with a With this in mind the user can select which variable the controller will put more emphasis on forcing to 0 The result of any stable full state feedback controller will be to drive all states to 0 qg 000 g 000 0 000 0 000 The above Q matrix is the simple design matrix you will be using for this controller In actuality the user may vary all the elements of the Q matrix to obtain the optimal controller but this is not necessary for the purpose of this lab Under the same directory open a model called s_SRV02_Inverted_Pendulum mdl This model is a simulation of the Linear model with a feedback law u kx The gain vector k is set by the LQR function Before beginning the simulation open a Matlab script file called Iteration_RotPen m Keep this script open as you will be executing it repetitively You may save amp run this file by pressing the F5 key Run this script and then proceed to the Simulink model and start the simulation You should have 3 scopes open that are displaying alpha a theta and the control signal u Every time the Iteration_RotPen m script runs you will see the new response in the simulation as well as two step responses 8 amp a in MATLAB figures If you right click on these figures and choose Characteristics gt Settling Time the settling time of the step response will be shown The default Q values are
6. able of 1 It is assumed that the student has successfully completed Experiment 0 of the SRV02 and is familiar in using WinCon to control the plant through Simulink It is also assumed that all the sensors and actuators are connected as per dictated in the SRVO2 User Manual and the Rotary Pendulum User Manual Page 2 Revision 01 3 Mathematical Model Figure 1 below depicts the Rotary inverted pendulum module coupled to the SRVO2 plant in the correct configuration The Module is attached to the SRVO2 load gear by two thumbscrews The Pendulum Arm is attached to the module body by a set screw The Inverted Pendulum experiment is a classical example of how the use of control may be employed to stabilize an inherently unstable system The Inverted pendulum is also an accurate model in the pitch and yaw of a rocket in flight and can be used as a benchmark for many control methodologies Figure 1 SRV02 w ROTPEN Module The following table is a list of the nomenclature used is the following illustrations and derivations Symbol Description Symbol Description L Length to Pendulum s Center ofh Distance of Pendulum Center mass of mass from ground m Mass of Pendulum Arm Jom Pendulum Inertia about its center of mass r Rotating Arm Length Vi Velocity of Pendulum Center of mass in the x direction 0 Servo load gear angle radians V Velocity of Pendulum Center Pendulum Arm Deflection radians
7. controller that will balance the inverted pendulum in the upright position The controller should also maintain the pendulum s stability upright while allowing the user to command varying setpoints of the servo angle The controller specifications for this lab are The controller must maintain system stability The controller must accurately place the servo angle theta at a given command while maintaining the pendulum in the upright position The pendulum angle a must not exceed 6 when the system is excited with a step input to the command The control signal Vm must be strictly bounded by 2 5 Volts The settling times of 0 amp a should not exceed 2 sec each for a step input 8 command of 1 rad 4 1 Partl Verification of the Linear Model The first part of this lab will be verifying our linear model by simulating the linear and non linear systems together and looking for the point of divergence This exercise will not only verify our linear model but will also establish a threshold to turn the controller off when a exceeds the threshold The first task upon entering the laboratory is to familiarize yourself with the system The pendulum deflection signal a should be connected to encoder channel 1 and the servomotor s position signal 8 should be connected to encoder channel 0 Analog Output channel 0 should be connected to the UPM Amplifier and from the amplifier to the input of the servomotor This system has o
8. d with a step input to the command The control signal Vm must be strictly bounded by 2 5 Volts The settling times of 0 amp a should not exceed 2 s each for a step input 98 command of 1 rad The first task upon entering the laboratory will be to simulate the non linear model of the rotary inverted pendulum and the above state space linear model The goal of this simulation is to determine the limits of the linear model what range of a does the linear model accurately describe the motion of the pendulum The following equations below describe the complete non linear system The linear system is derived by linearizing these two equations about a 0 The pre lab assignment is to determine a rough estimate of the range of a that the linear model will hold true NaN K p R m m ab bcos amp bsin c 0 G0 c amp bcos cx dsin 0 Page 7 Revision 01 4 In Lab Procedure The Rotary inverted pendulum is an ideal experiment when introducing important controls concepts such as disturbance rejection and non linear systems The intent of this laboratory is to first validate the use of the linear model through simulation and then proceed to design a closed loop full state feedback controller that will maintain the pendulum in its upright position The inverted pendulum has many practical applications in that it can be used to model many non linear systems The purpose of the lab is to design a state feedback
9. designed controller on the real plant in Part Ill of the lab you were asked to compile a Measurement Table Make sure this is included in your final report In Part Ill you implemented your controller on the physical plant Comment on the performance of your controller on the actual system as opposed to the simulated model Make sure to include your final controller gains and any re iterative calculations made if any Post Lab Questions 1 Having performed this lab by using the LQR technique what other control approaches would you consider for this system 2 The inverted pendulum is a classical example of a non minimum phase system In what way is this true Prove your argument with plots from the actual plant 3 By how much did your measurements of the actual controller deviate from those of the simulated model Give some quantitative measurements and comment on what you perceive as being the source of these discrepancies 4 Once you implemented your controller on the actual system you should have noticed the Limit Cycle phenomenon What ways would there be of incorporating this into your model so that it would show up in the simulations Page 14 Revision 01
10. f the pendulum setup you are using is not in the same configuration as set in the setup file you should notify the lab technician or T A to correct the error Now that the pendulum arm is in the correct starting position as set in the setup file you are ready to start the controller Note Figure 6 shows the PENDULUM_START in the UP position Notice how the pendulum ar is sitting flush on the right stopper Figure 5 shows the PENDULUM_START in the DOWN position Notice how the pendulum arm is fastened to the tip of the shaft Start the controller through WinCon The controller is now running but the plant is not moving as you must first enable the system by having the pendulum arm cross the a 0 threshold With one finger on the tip of the pendulum slowly move the pendulum toward the upright position until you feel the system begin to work This controller was designed to remain dormant until the pendulum arm has crossed the upright position and enabled the controller Your controller should now be running WARNING If at any point the system is not behaving as expected make sure to immediately press STOP on the WinCon server If at any time you hear a high frequency hum from the system this is an indication that the gains are too high and you need to re calculate your controller Through the WinCon server open a plot of 3 signals Measured Alpha Measured Theta amp Control Signal You should be noticing the postion of the se
11. h between 1 4 amp 1 6 seconds Here you will see the two curves begin to diverge It seems that the linear model quite accurately describes the system for the first 25 and then begins to diverge from the actual motion We will therefore set the Alpha_Threshold variable to 25 It should be noted that a range of 25 is quite acceptable for this application as the controller you will be designing will try to maintain the pendulum in the upright position a 0 4 2 Part Il Designing amp Simulating the Controller Now that we have verified the linear model to be accurate in the range of motion we are expecting you will now use this linear model to design a full state feedback controller to maintain the pendulum in the upright position and to meet the controller specifications mentioned above The method of calculating the feedback gains will be the LQR function in MATLAB s control systems toolbox The MATLAB LQR function returns a set of calculated gains based on the system matrices A amp B and the design matrices Q amp R In this section of the Lab you will begin the iterative design process by varying Q and taking notice on the effect those changes have on the simulated system response Page 9 Revision 01 For the purpose of this lab we will fix the Q matrix to only be comprised of 2 elements q1 q2 and we will hold R at 1 The reason behind limiting the design to only 2 variables is to investigate the role of emphasis in an
12. i SRV02 Series 9 Rotary Experiment 7 MUA NS ER Rotary Inverted Pendulum Student Handout SRV02 Series Q Rotary Experiment 7 GUANSER Rotary Inverted Pendulum Student Handout 1 Objectives The objective in this experiment is to design a state feedback controller for the rotary inverted pendulum module using the LQR technique The controller will maintain the pendulum in the inverted upright position and should be robust in order to maintain its stability in the case of a disturbance Upon completion of the exercise you should have have experience in the following How to mathematically model the rotary inverted pendulum system To linearize the model about an equilibrium point To use the LQR method in designing a state feedback controller To design and simulate a WinCon controller for the system To use feedback to stabilize an unstable system 2 System Requirements To complete this Lab the following hardware is required 1 Quanser UPM 2405 1503 Power Module or equivalent 1 Quanser MultiQ PCI MQ3 or equivalent data acquisition card 1 Quanser SRVO2 servo plant 1 Quanser ROTPEN Rotary Pendulum Module 1 PC equipped with the required software as stated in the WinCon user manual The required configuration of this experiment is the SRV02 in the High Gear configuration along with a ROTPEN Rotary Pendulum Module as well as a UPM 2405 1503 power module and a suggested gain c
13. ne input Vm and two outputs 8 amp a You are now ready to begin the lab Launch MATLAB from the computer connected to the system Under the SRV02_Exp7_Inverted Pendulum directory begin by running the file by the name Setup_SRV02_Exp7 m This MATLAB script file will setup all the specific system parameters and will set the system state space matrices A B C amp D You are now ready verify the linear model Under the same directory open a Simulink model called s SRV0O2 Non_Linear_ Pendulum mdl This model is a simulation of the complete non linear pendulum system Page 8 Revision 01 as well as the linearized state space representation of the system The purpose of this simulation is to investigate the validity of the linear model and also to determine the range of a where the linear model correctly describe the system Start the simulation You should have 2 scopes open that are displaying alpha a and theta 8 To put the plots in perspective a was given an initial condition slightly greater than O and thus allowed to fall Figure 4 Below is a plot of the linear and non_linear simulations of a Alpha _ 5 x bBlog p ABB oa t Figure 4 Simulation of Non Linear amp Linear Models As you can seen from the above simulation the linear model correctly depicts the motion of the pendulum for the first 1 5 seconds and then begins to break down In the simulation plot of alpha zoom into grap
14. otential energy in the system is gravity V P E mgh mgLcose 3 4 Pendulum Kinetic Energy The Kinetic Energies in the system arise from the moving hub the velocity of the point mass in the x direction the velocity of the point mass in the y direction and the rotating pendulum about its center of mass T K E yy tK Ey tK Ey KE 3 5 Pendulum Note Since we have modeled the pendulum as a point mass at its center of mass the total kinetic energy of the pendulum is the kinetic energy of the point mass plus the kinetic energy of the pendulum rotating about its center of mass The moment of inertia of a rod about its center of mass is J l yR cm 12 since we ve defined L to be half the pendulum length then R in this case would be equal to 2L Therefore the moment of inertia of the pendulum about its center of mass is l 2 1 2 l 2 P22 MRE tmon mi 3 6 om 12 pias 3 6 Finally our complete kinetic energy T can be written as T 8 m r L cosa a E m L sinala PES a 3 7 After expanding equation 3 7 and collecting terms we can formulate the Lagrangian L T V 5J 0 5 m1 mLr eoset 8 5 m r0 mgLcos 3 8 Page 5 Revision 01 Our 2 generalized co ordinates are 8 and a We therefore have 2 equations fL L ai SiN oa 5a oo Pa 3 9 STALAN EL ie al 3 10 bt 6a Sa Solving Equations 3 9 amp 3 10 and linearizing about a 0 we are left with BO 3 11 output eq
15. rly show how the requirements have been met Before closing the Iteration_RotPen m file make sure you run it one last time with your final design parameters q amp q2 such that the gain vector k is set accordingly You may now stop the simulation 4 3 Part Ill Implementing the Controller After verifying the calculated controller gains it is time to implement the controllers on the actual system In the same working directory open a Simulink model called q_SRV02_Inverted_Pendulum mdl This model has the I O connection blocks linking to the physical plant as well as a simulated block to compare real and simulated results You may now proceed to Build the controller through the WinCon menu Before starting the controller make sure that the pendulum rod is in its correct staring position The starting position of the pendulum should match the setting given in the experiment setup file Setup _SRV0O2_Exp7 m Open this file and scroll to the following section HHHHHHHHHHHH USER DEFINED PENDULUM CONFIGURATION Pendulum Start Position set to UP or DOWN PENDULUM_START UP HHHH END OF USER DEFINED PENDULUM CONFIGURATION Figure 6 PENDULUM_START UP Figure 5 PENDULUM_START DOWN Page 11 Revision 01 Referring to Figure 5 amp Figure 6 please make sure that the pendulum arm is in the correct starting location as is set in the Setup _SRV02_Exp7 m file I
16. rvo Theta is oscillating This is referred to as a Limit Cycle and is cause by the static friction in the system A more thorough investigation of this phenomenon is out of the scope of this laboratory You must now ensure that you controller meets the required specifications Create a table similar to the Measurement Table below Switch back to the Simulink diagram and enter a Theta Setpoint of 28 28 is approximately equal to 0 5 radians Due to the Limit Cycle it is necessary to gather measurements of 5 step inputs and compute the average As your controller is sending a step input into the system you should be filling out your Measurement Table It is recommend to change the buffer of the WinCon plots to 10 seconds as to get 2 full cycles of data per plot Sample Alpha Range Control Signal u Range 1 4 5 lt a lt 5 5 2V lt u lt 2V 5 4 7 lt a lt 5 3 2 2V lt u lt 2 2V Average 4 63 lt a lt 5 4 2 05V lt u lt 2 1V Table 2 Measurement Table Sample Entries Page 12 Revision 01 E scope q_SR O2_Inverted_Pendulum Measured Alpha deg File Edit Update Axis Window Background Colour Text Colour Text Font 2 Figure 8 Control Signal Capture Using the average as an indication of your overall controller performance does your controller still meet the system requirements Due to the Limit Cycle the Alpha amp
Download Pdf Manuals
Related Search