SYSTEM IDENTIFICATION AND POSITION CONTROL OF PNEUMATIC ACTUATOR USING EMBEDDED SYSTEM TAHA MOHAMMED AHMED SADEQ

Transcription

1 SYSTEM IDENTIFICATION AND POSITION CONTROL OF PNEUMATIC ACTUATOR USING EMBEDDED SYSTEM TAHA MOHAMMED AHMED SADEQ A project report submitted in partial fulfilment of the requirements for the award of the degree of Master of Engineering (Electrical-Mechatronics &Automatic Control) Faculty of Electrical Engineering Universiti Teknologi Malaysia JANUARY 2015

2 ^pec/^/{y JeJz'c%?eJ wy ^^re^^^, wy _/^wz7y wy _/rz'e^j^ _/or ^Aezr ^^^ze^ce, ^Mppor^, ^r^yer^, e^commgewe^ ^/e^^z^g^

3 iv ACKNOWLEDGEMENT First and foremost, I would like to express my thanks to Allah Subhanho wa ta'ala who is helped and guided me to finish this study. Second, I would like then to express my sincere appreciation to my supervisor, Dr. Ahmad Athif Mohd Faudzi for his continuous guidance, professional advices, support, motivation, and encouragement to complete this study. Also, I would like to extend my thanks to all my colleagues, friends, and my research group members for support and help. Moreover, I would like to express my gratitude to all my professors, lecturers and staff in Universiti Teknologi Malaysia for their continuous guidance and help during my study. Last but not least, I would like to thanks the Ministry of Higher Education and Scientific Research-YEMEN for supporting my study.

4 v ABSTRACT This project report presents the ability of using embedded system to control non-linear systems like pneumatic actuators. The pneumatic actuators are commonly used in industrial applications because they have many practical advantages such as high power to weight ratio and cheap compared to other actuators. Embedded system STM32F4DISCOVERY was used in this project to interface the pneumatic actuator with MATLAB Simulink instead of using the conventional Data Acquisition card (DAQ) to reduce the cost and size. The embedded system was also used as a controller to execute the control algorithms. The model of the pneumatic actuator was identified by using system identification technique. Based on this model, the Proportional-Integral-Derivative (PID) and the Proportional-Derivative Fuzzy Logic (PD-Fuzzy) controllers were designed to control the position of the pneumatic actuator. Waijung Blockset in MATLAB Simulink was used to estimate the model of the pneumatic actuator and implement these controllers inside the embedded system STM32F4DISCOVERY. A Pneumatic Actuated Ball and Beam System (PABBS) is an application of controlling the position of the pneumatic actuator. The developed model of PABBS was used to design three controllers for this plant based on the controllers of the pneumatic actuator. The Cascade PID, the PD-Fuzzy with Gain Feedback and the PD-Fuzzy with PID controllers were designed to control the ball at the desired position. For the position control of the pneumatic actuators, the simulation and experimental results of the PID and PD-Fuzzy controllers were presented. The PD-Fuzzy offers better control compared to other controllers in terms of stability and robustness for the pneumatic actuator. On the other hand, PD-Fuzzy with PID controllers gave the best control response in the simulation compared to the others controllers for the PABBS application.

5 vi ABSTRAK Tesis ini membentangkan keupayaan menggunakan sistem terbenam untuk mengawal sistem tidak linear seperti penggerak pneumatik. Penggerak pneumatik biasanya digunakan dalam aplikasi industri kerana ia mempunyai banyak kelebihan praktikal seperti nisbah tenaga kepada berat yang tinggi selain murah berbanding dengan penggerak lain. Sistem terbenam STM32F4DISCOVERY telah digunakan dalam projek ini untuk menghubungkan penggerak pneumatik dengan MATLAB Simulink disebalik menggunakan DAQ konvensional (Data kad Perolehan) untuk mengurangkan kos dan saiz. Sistem tertanam juga digunakan sebagai pengawal untuk melaksanakan algoritma kawalan. Model penggerak pneumatik telah dikenal pasti dengan menggunakan teknik Pengenalpastian Sistem (SI). Berdasarkan model ini, pengawal PID dan PD-Fuzzy telah direka untuk mengawal kedudukan penggerak pneumatik. Waijung Blockset dalam MATLAB Simulink telah digunakan untuk menganggar model penggerak pneumatik dan melaksanakan pengawal ini di dalam sistem STM32F4DISCOVERY terbenam. Sistem bebola dan palang pneumatic (PABBS) adalah sebuah aplikasi bertujuan mengawal kedudukan penggerak pneumatik. Model yang dibangunkan daripada PABBS telah digunakan untuk mereka bentuk tiga pengawal untuk pelan ini berdasarkan pengawal penggerak pneumatik. PID lata, PD-Fuzzy dengan maklumbalas dan PD-Fuzzy dengan pengawal PID yang telah direka untuk mengawal bola pada kedudukan yang dikehendaki. Bagi kawalan kedudukan daripada penggerak pneumatik, simulasi dan keputusan eksperimen daripada PID dan pengawal PD-Fuzzy telah dibentangkan. Kawalan PD-Fuzzy adalah lebih baik berbanding dengan pengawal lain dari segi kestabilan dan kemantapan untuk penggerak pneumatik. Sebaliknya, PD-Fuzzy dengan pengawal PID memberi tindak balas kawalan terbaik dalam simulasi berbanding dengan yang pengawal lain untuk aplikasi PABBS.

6 vii TABLE OF CONTENTS CHAPTER TITLE PAGE DECLARATION DEDECATION ACKNOWLEGEMENT ABSTRACT ABSTRAK TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES LIST OF ABBREVIATIONS ii iii iv v vi vii x xi xvi 1 INTRODUCTION Background Problem Statement Research Objective Scope of the Work 4 2 LITERATURE REVIEW Intelligent Pneumatic Actuator (IPA) System Identification of Pneumatic Actuator Ball and Beam System using Electrical Actuator Ball and Beam System using Pneumatic Actuator Control the Position of Pneumatic Actuator 13

7 viii 3 METHODOLOGY The Embedded System STM32F4DISCOVERY Waijung Blockset Interface setup System Identification Techniques Parametric Models Modeling of Pneumatic Actuated Ball and Beam Controllers Design for Pneumatic Actuator PID Controller PD-Fuzzy Controllers Controllers Design for PABBS Cascade PID Controllers Design PD-Fuzzy with Gain Feedback Controller Design PD-Fuzzy with PID Controller Design Experimental Setup 31 4 RESULTS AND DISCUSSION System Identification for the Pneumatic Actuator Position Control for Pneumatic Actuator Simulation results Experimental results Comparison between Simulation and Experiment for Position Control Position Control for Pneumatic Actuated Ball and Beam System The Simulation result of Cascade PID Controllers The Simulation result of Fuzzy with Gain Feedback Controller The Simulation result of PD-Fuzzy with PID Controllers Comparison between PABBS Controllers 50

8 ix 5 CONCLUSIONS AND RECOMMENDATIONS Conclusion Future W ork 53 REFERENCES 54

9 x LIST OF TABLES TABLE NO. TITLE PAGE 2.1 Specification of the pneumatic actuator 6 system 3.1 PABBS dimensions System Parameters for PABBS The criteria of fuzzy controller Input and output terminals configurations 32 on STM32F4DISCOVERY 4.1 Comparisons for step response analysis 40 results for position -Simulation 4.2 Comparisons for step response analysis 43 results for position - experimental 4.3 Comparisons for step response analysis 50 results for PABBS -Simulation

10 xi LIST OF FIGURES FIGURE NO. TITLE PAGE 1.1 Types of Ball and beam system Intelligent actuator elements Schematic diagram valve connection of 7 the pneumatic cylinder 2.3 interface method between (IPA) and 8 MATLAB 2.4 Quanser Ball and beam module Top view of the dual motors ball and 10 beam system 2.6 Z '.S ltum and J.Petric Ball and beam 11 module 2.7 equipment's for PABBS System project Flowchart of Methodology STM32F4DISCOVERY embedded 18 system 3.3 library of Waijung Blockset UART connection 20

11 xii 3.5 STM32F4DISCOVERY, USB Converter- 21 N and amg F4 Connect 2 boards 3.6 System Identification Flow Chart PABBS System Free body diagram of PABBS System The architecture of Pneumatic actuator 26 controllers design 3.10 The architecture o f PD-Fuzzy controller The architecture of PABBS controllers 29 design 3.12 The architecture of feedback controller 30 design 3.13 Experimental parts setup The Simulink files for SI The Simulink block for PWM Input and output data for SI SI toolbox after insert input and output 36 data 4.5 The best fit percentage for the estimated 37 model 4.6 Simulation blocks for PID and PD-Fuzzy 38 controllers for position control 4.7 Structure blocks of the position 38 controllers

12 Simulation responses for multistep 39 position control Experimental Targets Blocks for control 41 the position Waijung track build process window 42 Experimental responses for multistep 43 position control (a) PID (b) PD-Fuzzy Comparison between simulation and 44 experimental results Simulation blocks of Cascade PID 45 controllers for PABBS Structure blocks for subsystem blocks of 46 PABBS Simulation responses of Cascade PID 47 controllers for PABBS Simulation blocks of PD-Fuzzy with Gain 48 Feedback controllers for PABBS Simulation responses of PD-Fuzzy with 48 Gain Feedback controllers for PABBS Simulation blocks of PD-Fuzzy with PID 49 controllers for PABBS Simulation responses of PD-Fuzzy with 50 PID controllers for PABBS Comparison between simulation 51 responses for PABBS controllers

13 xiv LIST OF ABBREVIATIONS ARX - Auto-Regressive with Exogenous ARMAX - Auto-Regressive M oving Average with Exogenous DAQ - Data Acquisition FIS - Fuzzy Inference System FLC - Fuzzy Logic Control I/O - Input/ Output MF - Membership Function PABBS - Pneumatic Actuated Ball and Beam System PD - Proportional-Derivative PI - Proportional-Integral PID - Proportional-Integral-Derivative PWM - Pulse Width Modulation SI - System Identification SISO - Single Input Single Output TF - Transfer Function

14 CHAPTER 1 INTRODUCTION 1.1 Background Pneumatic actuators are often used to convert the energy of compressed air into mechanical energy. They are widely used in robotics, medical applications and automation applications because they offer many useful advantages. Also, electric motors are used as actuators in various applications like robotics and automation applications because they are easy to control and interface precisely. But, electrical actuators have several disadvantages such as low power to weight ratio, high current related with its load and heavy. The hydraulic actuators are not ecological, need hydraulic oil and their pumps need return lines. On the other hand, the pneumatic actuators are clean, economy, light, fast, have a great power to weight ratio and do not need return lines. However, the pneumatic actuators have highly nonlinear characteristics due to the air compressibility behavior and internal friction. Because of these conditions, there are certain difficulties in pneumatics cylinder control design [1]. One of these difficulties is the usage of pneumatic actuators cannot be controlled precisely. Furthermore, pneumatic actuators have nonlinear, high-order dynamics and parametric uncertainties due to friction and air compressibility [2].

15 2 Ball and beam system is one of the most common laboratory systems and it is classified under tracking control. This prototype is used for education purposes and experimental verifications of applied control algorithms. This system depends on controlling the position of a ball rolling on the top of a long beam by adjusting the angle of the beam. The dynamics of this system are complex and unstable open loop system. There are many prototypes of ball and beam system that use D C motor to control the angle of the beam. Recently, the researchers have been trying to design and build a new prototype of ball and beam system that uses pneumatic actuator to control the angle of the beam. This prototype has many non-linear characteristics that cause difficulties in controlling this system. Figure 1.1 represents the common actuators that use in ball and beam system. Embedded systems are a new technique that came with the revolution of technology and can be seen everywhere in our life. An embedded system is a specialpurpose computer system designed to perform some decided functions. IEE defines the embedded system as a device used to control, monitor or assist the operation of equipment, machinery or plant [3]. The embedded system consists of microprocessor, RAM, ROM, and I/O peripheral. All computing systems have constraints on design metrics unlike those in embedded system. A design metric is a measure of an implementation's features, such as cost, size, performance, and power. Embedded systems are cheap, have small size, fixed in one chip, active suitable performance to process data in real-time, and have low power consumption. Figure 1.1 Types of Ball and beam system

16 3 1.2 Problem Statement Due to the several advantages of pneumatic actuators, they are used in many applications. On the other hand, it is difficult to control the position of pneumatic actuator because of its non-linear characteristics. The prototype of the Pneumatic Actuated Ball and Beam System (PABBS) is a new application of controlling the position of the pneumatic actuator as in [4]. The significance of this prototype is due to the challenge of controlling the position of the ball along the beam by controlling the beam angle. In PABBS, the sensors and the actuators are interfaced with computer via National Instrumentation Data Acquisition card (NI-DAQ), to control the system in real-time using MATLAB Simulink blocks. The designer uses the personal computer to apply the control algorithm via the interface between the computer and the system, which leads to more power consumption and cost. In addition, the big size of personal computer causes another drawback. 1.3 Research Objectives * To develop the model of Pneumatic Actuator by using System Identification (SI) technique. * To design PID and PD-Fuzzy controllers for controlling the position of Pneumatic Actuator in MATLAB and implement them inside embedded system (STM32F4DISCOVERY). * To design three different controllers to control Ball and Beam system in MATLAB Simulink.

17 4 1.4 Scope of Work * Use system identification to estimate the linear position model of Pneumatic Actuator and use this model to design PID and PD-Fuzzy controllers to control the position of the pneumatic actuator. * Implement the PID and PD-Fuzzy controllers inside the flash memory of the embedded system (STM32F4DISCOVERY) board. * Build the Cascade model of Pneumatic Actuated Ball and Beam system and design the controllers for this system based on the position controllers of the pneumatic actuator by using MATLAB Simulink blocks. * The interfaced with MATLAB Simulink in real-time just to measure the response.

18 REFERENCES [1] Ramos-Arreguin, J. M., Pedraza-Ortega, J. C., Gorrostieta-Hurtado, E., Romero-Troncoso, R. J., Vargas-Soto, J. E., & Hernandez-Hernandez, F. (2008). Pneumatic Fuzzy Controller Simulation vs Practical Results for Flexible Manipulator. ^M?ow%%oM %MJ ^o^o%c.y, 5, [2] Xu, B., Hayakawa, Y., & Pandian, S. R. (2005, November). A sliding mode fuzzy force tracking controller for pneumatic cylinders. 7M <Sywpos;'MW om E/Mz'J Pow er (Vol. 2005). [3] Xu, R. Z., Guo, W., & Li, C. X. (2006, May). A New Type of Embedded Autolevelling Control System based on ARM Microcontroller for Carding Machine. In TMJMsr'%/ E/ec?roMzcs %MJ ^pp/z'c%%oms 200d 7EEE CoM/ereMce om (pp. 1-6). IEEE. [4] Muhammad Asyraf Bin Azman, PoszWoM com?ro/ MszMg EMzzy-^^seJ om com?ro//er /o r pmemw%%c servo cy/zmjer zmm // ^MJ ^e^m ^pp/zc^?zom, Master 2014 Universiti Teknologi Malaysia. [5] A. A. M. Faudzi. Deve/opweM? o / 7M?e//zgeM? PMeMW^^zc ^c?m^?ors ^MJ TAez'r ^pp/zc^?zoms?o PAyszc^/ NMW^M-MecAzMe 7M?er%c%oM ^ys?ew. PhD. The Graduate School of Natural Science and Technology, Okayama University; [6] Faudzi, A. A. M., Suzumori, K., & Wakimoto, S. (2010). Development of an intelligent chair tool system applying new intelligent pneumatic actuators. ^Jv%MceJ ^o^o^'cs, 24(10), [7] Faudzi, A. M., Suzumori, K., & Wakimoto, S. (2009, July). Design and control of new intelligent pneumatic cylinder for intelligent chair tool application. In ^Jv%MceJ 7M?e//zgeM? MecA%?roMz'cs T E E E/^M E 7M?erM^?z'oM^/ CoM/ereMce om (pp ). IEEE. [8] Rahmat, M. F., Salim, S. N. S., Sunar, N. H., Faudzi, A. A. M., Hilmi, Z., & Huda, I. K. (2012). Identification and non-linear control strategy for industrial pneumatic actuator. 7M?erM%%oM%/ JoMrM%/ o / PAysz'c%/ ^czemces, 7(17), [9] K. Osman, A. A. M. Faudzi, M. F. Rahmat, N. D. Mustafa, M. A. Azman and K. Suzumori (2012). System Identification model for an Intelligent

19 55 Pneumatic Actuator (IPA) system. 7^eff/geM? Ro^o?s a ^ J Sys^ews (7ROS), EEERSJ 7^^er^a^/o^af Co^/ere^ce o^ Oct [10] Peng, X., Li, L., & Zhi, T. (2010, July). "Human simulated intelligent control for ball and beam system". 7^ Co^^rof Co^/ere^ce (CCC), P^A CA/^ese (pp. P6-700). 7EEE. [11] Gunasekaran, M., & Potluri, R. (2008, December). "Cooperative control of a dual-motor ball and beam system". 7^ 7^J/a Co^/ere^ce, ND7CO# ^^^Maf 7EEE (Vol. 2, pp ). [12] SSitum, Z., & Petrie, J. (2008). "A pneumatically actuated ball and beam system". 7^^er^a^/o^af JoMr^af / MecAa^/caf E^gmeermg EJMca^/o^, 36(3), [13] Chillari, S., Guccione, S., & Muscato, G. (2001). An experimental comparison between several pneumatic position control methods. In Decision and Control, ProceeJmgs /?Ae 40?A 7EEE Co^/ere^ce o^ (Fof. 2, pp ;. 7EEE. [14] STMicroelectronics, September 2011, [15] Aimagin Co,Ltd, 2014, [16] National Instruments, 2014, [17] Shojaefard, M. H., Ghaffarpour, M., & Noorpoor, A. R. (2008). Thermal contact analysis using identification method. Nea? Tm^s/er E^gmeermg, 29(1), [18] C. Wenzhuo, S. Xiaomei and X. Yonghui (2012). Modeling and modulation of nonlinear ball-beam system controller based on matlab. 9?A 7^erMa%o^af Co^/ere^ce o^ FMzzy Sys^ews a ^ J X^^owfeJge D/scovery (FSXD) May [19] S.-K. Oh, H.-J. Jang and W. Pedrycz (2011). Optimized fuzzy PD cascade controller: A comparative analysis and design. S/wMfa?o MoJeffmg Prac^/ce an J TAeory. 19 (1), [20] Rozali, S. M., Rahmat, M. F., Wahab, N. A., & Ghazali, R. (2010, December). PID controller design for an industrial hydraulic actuator with servo system. In ResearcA a ^ J DevefopweM? (SCOReD), EEE S?MJeM? Co^/ere^ce o^ (pp ). IEEE.

20 56 [21] Nhivekar, G. S., Nirmale, S. S., & Mudholker, R. R. (2011). Implementation of fuzzy logic control algorithm in embedded microcomputers for dedicated application. 7^er^%%o^%f JoMr^^f o / F^gmeermg, ^cze^ce J(4). TecMofogy, [22] Zhuang, M., & Atherton, D. P. (1994, March). Optimum cascade PID controller design for SISO systems. In Co^^rof, Co^^rof'94. 7^er^%%o^%f Co^/ere^ce o^ (Vol. 1, pp ). IET. [23] Azman, M. A., Faudzi, A. A. M., Osman, K., & Natarajan, E. (2014). Integrating Servo-Pneumatic Actuator with Ball Beam System based on Intelligent Position Control. JMr^%f TeA%ofog;', 69(3). [24] Krastev, K. K., Gotsov, T. A., & Vladimirova, A. Embedded System for Measurements and Data Acquisition with Web Interface Based on ARM Processors, JoMr^^f 0/*Ffec^ro^zc^, [25] Keshmiri, M., Jahromi, A. F., Mohebbi, A., Amoozgar, M. H., & Xie, W. F. (2012). Modeling and control of ball and beam system using model based and non-model based control approaches. 7^er^%%o^%f JoMr^^f o^.sw%r?.ye^mg z^^effzge^^ 3(7), 74-JJ. [26] Choi, G. S., Lee, H. K., & Choi, G. H. (1998). A study on tracking position control of pneumatic actuators using neural network. 7^ T ^M ^r^f Ffec^ro^zc^ <Sbc;'e(y, 799S. IECON'98. Proceedings of the 24th Annual Conference of the IEEE (Vol. 3, pp ). IEEE. [27] Khan, S., Abdulazeez, S. F., Adetunji, L. W., Alam, A. Z., & Salami, M. J. E. (2008). Design and implementation of an optimal fuzzy logic controller using genetic algorithm. o / co^m ^er ^cze^ce, 4(70), 799 [28] S. N. S. S. M.F. Rahmat, Ahmad Athif Mohd Faudzi, Zool Hilmi Ismail, S I. Samsudin, N.H.Sunar and K.Jusoff (2011). Non-linear Modeling and Cascade Control of an Industrial Pneumatic Actuator System. JoMr^^f o / F^^z'c ^^pfzej ^cze^ce. J (S),