A dissertation submitted in fulfilment of the requirement for the degree of Master of Science (Mechatronics Engineering)

ii

ADAPTIVE BACKSTEPPING CONTROL OF PNEUMATIC ANTHROPOMORPHIC HAND

BY

MOHANNAD K. H. FARAG

A dissertation submitted in fulfilment of the requirement for the degree of Master of Science (Mechatronics Engineering)

Kulliyyah of Engineering

International Islamic University Malaysia

December 2016

ii

ABSTRACT

This research presents a nonlinear adaptive backstepping strategy for control of a pneumatic anthropomorphic robotic hand. An anthropomorphic hand with three fingers has been developed in this work. The fingers are driven by tendons and actuated by human muscle-like actuators known as Pneumatic Artificial Muscle (PAM). The high nonlinear dynamics of these actuators and the inherent hysteresis in their behaviour lead to the modelling and control problems that cause a lack of robustness in the hand's performance. The robotic finger and the PAM actuator have been mathematically modelled as a nonlinear second order system based on an empirical approach. An adaptive backstepping controller has been designed in two design steps based on the nonlinear second order system model for position control of the pneumatic anthropomorphic hand. In the design procedure the estimator of the system uncertainty is incorporated to the proposed control law which is extended for grasping objects with changing weights using a slip detection strategy. In addition, a cascade control system is developed by combining a conventional PID control, as the inner pressure control loop, with the adaptive backstepping control as the outer position control loop.

Simulation and experimental test have been conducted using an experiment setup to evaluate the performance of the designed controller. Based on both simulation and experimental results, the adaptive backstepping position controller is capable to compensate the uncertain coulomb friction force of PAM actuator achieving the desired angular trajectory with RMSE of hysteresis behaviour in range of 0.09^o - 0.18^o and RMSE of angular position control in range of 0.05^o - 0.11^o. The cascade controller has shown a stable supply of pressurized air with average settling time of 0.38 s - 0.57 s. In terms of force control, the robotic hand is able to maintain grasping objects when their weight is increased up to 500 g by detecting the slip signal generated by the force sensor.

Therefore, based on the obtained results, the controller is capable of tracking the desired position accurately and the pneumatic anthropomorphic hand is able to prevent the object from dropping when its weight is increased. For future researches, the adaptive backstepping position controller can be used to overcome other uncertain parameters such as the viscous friction. The further pneumatic hand can also be improved by increasing the number of the robotic fingers and DOFs to improve its manipulation and grasping ability.

iii

ثحبلا ةصلاخ

ABSTRACT IN ARABIC

" ـلا ةقيرط ثحبلا اذه مدقي backstepping

"

مكحتلل يطخلالا ةيتوبور ديب فيكتلما

ةيئاوه ديب ًةهيبش ةمسمج

ناسنلإا ريوطت تم ثيح عباصأ ةثلاثب ةيتوبور دي

هبشت ةيئاوه تلاغشبم ةطوبرم رتاوأ مادختسبا عباصلأا هذه كَّر حتُ .

" ـب ىمست ناسنلإا ةلضع

" ـل ًاراصتخا " PAM Pneumatic Artificial Muscle

."

في ةيلاعلا ةيطخلالا

ؤطابتلا ةرهاظو تلاغشلما هذه ةيكيمانيد "

hysteresis

"

تلاغشلما هذه ةجذنم في لكاشم لىإ نيادؤي اهكولسل ةمزلالما

.ةيتوبورلا يديلأا ءادأ ةيقوثو في ًاديدش ًاصقن لياتلبا ببسي امم ابه مكحتلاو ةيئاولها تلاغشلماو ةيتوبورلا عباصلأا تجِذحنم

ًايضيار لما مكحتلما ميمصت تم .ةجذمنلا في ةيبيرجتلا ةقيرطلا ىلع ًءانب ةيناثلا ةجردلا نم ةيطخ يرغ ةلادك ت

فيك

backstepping ديلبا يعضولما مكحتلل ةيناثلا ةجردلا نم يطلخا يرغ ماظنلا جذونم ىلع ءانب ميمصت تيلحرم للاخ

هذه ميمصتلا ةقيرط في .ةيتوبورلا مدع لماعم نممخ جمد تم

حترقلما مكحتلا نوناق في ماظنلا ينيعت يرخلأا عيسوت تم ثيح

كسمتل ديلبا مكحتلا لمشيل جأ

با كلذو ةفلتمخ نازوأ تاذ ماس مادختس

أدبي امدنع دلوتت تيلا ةراشلإا فشك ةقيرط

سلجا .هنزو ةدياز ةجيتن قلازنلابا م كلذ لىإ ةفاضإ

، فيكتلما مكحتلما لولأا ،ينمكحتم نم نيبم لياتتم مكحتم ميمصت تم

backstepping –

عضولمبا مكحتلل ةيجراخ ةقلحك –

مكحتم نياثلاو PID

ةيلخاد ةقلحك يئاوه طغض مظنم عم

ابتخا ذيفنت تم .طغضلبا مكحتلل ر

ةاكامح رابتخاو تابثإو مييقتل بييرتج تم ثيح مكحتلا ماظن لمع ةحص

دادعإ رتج بيي

وبورلا ديلل ةيت

ةجيتنلبا .

، نإف فيكتلما مكحتلما backstepping

رهظأ لا نكاسلا كاكتحلاا ينمتخ ىلع ةردق ًاضوعم

ةرهاظ كلذب ـلا

ؤطابت "

hysteresis عضولمبا مكحتلا لامج في "

لوصولا تم ثيبح أطبخ راسملل

RMS ؤطابتلا ةرهاظل

o

ينب 0.18

o

– أطخو 0.09

ـب ردقي عضولمبا مكحتلل RMS ا

لالمج 0.11

–

0.05 . لما رهظأ مكحت PID

ديوزت في ًارارقتسا للاخ طوغضلما ءاولهبا ةيئاولها تلاغشلما

طسوتم ينب حاوتري رارقتسا نمز 0.38

ةينثا و 0.57 ةينثا .

مكحتلا لامج في ب

مسلجبا ديلا كاسمإ ةوق لع ظفاح مكحتلا ماظن نإف

هنزو ةدياز دنع ًاكوسمم مسلجا ى إ

دح لى 500

مارغ با شتك فا مسلجا قلازنا ةراشإ ةوقلا تاساسح نم ةدلولما

. ةصلختسلما جئاتنلا ىلع ءانب مكحتلا ماظن نإف كلذبو

في ًاديج ًءادأ رهظأ لكك وغرلما عضولما ءافتقاو ةعباتم

سلجا طوقس عنم ىلع ةردقلما اله ةيئاولها ةيتوبورلا ديلاو ب دنع م

هنزو ةدياز .

ةيلبقتسم تاسارد لجأ نم نإف

مكحتلا ماظن فيكتلما

backstepping تخ في همادختسا نكيم

ينم

تلاماعم ىرخأ ةنيعم يرغ

جزللا كاكتحلااك

ينسحتل اهتيرح تاجردو ديلا عباصأ ددع ةدياز نكيم امك في ديلا ةردق

.ةروانلماو كاسملإا

iv

APPROVAL PAGE

I certify that I have supervised and read this study and that in my opinion, it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Master of Science (Mechatronics Engineering).

………..

Norsinnira Zainul Azlan Supervisor

………..

Salmiah Ahmad Co-Supervisor

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Master of Science (Mechatronics Engineering).

………..

Md. Raisuddin Khan ………….

Internal Examiner

………..

Ali Sophian ………….

Internal Examiner

This dissertation was submitted to the Department of Mechatronics and is accepted as a fulfilment of the requirement for the degree of Master of Science (Mechatronics Engineering).

………..

Tanveer Saleh Head, Department of Mechatronics Engineering

This dissertation was submitted to the Kulliyyah of Engineering and is accepted as a fulfilment of the requirement for the degree of Master of Science (Mechatronics Engineering).

………..

Md Noor Bin Salleh

Dean, Kulliyyah of Engineering

v

DECLARATION

I hereby declare that this dissertation is the result of my own investigations, except where otherwise stated. I also declare that it has not been previously or concurrently submitted as a whole for any other degrees at IIUM or other institutions.

Mohannad K. H. Farag

Signature ... Date ...

vi

COPYRIGHT PAGE

INTERNATIONAL ISLAMIC UNIVERSITY MALAYSIA

DECLARATION OF COPYRIGHT AND AFFIRMATION OF FAIR USE OF UNPUBLISHED RESEARCH

THE IMPACT OF MOBILE INTERFACE DESIGN ON INFORMATION QUALITY OF M-GOVERNMENT SITES

I declare that the copyright holders of this dissertation are jointly owned by the student and IIUM.

Copyright © 2016 Mohannad K. H. Farag and International Islamic University Malaysia. All rights reserved.

No part of this unpublished research may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without prior written permission of the copyright holder except as provided below

1. Any material contained in or derived from this unpublished research may be used by others in their writing with due acknowledgement.

2. IIUM or its library will have the right to make and transmit copies (print or electronic) for institutional and academic purposes.

3. The IIUM library will have the right to make, store in a retrieved system and supply copies of this unpublished research if requested by other universities and research libraries.

By signing this form, I acknowledged that I have read and understand the IIUM Intellectual Property Right and Commercialization policy.

Affirmed by Mohannad K. H. Farag

……..……….. ………..

Signature Date

vii

ACKNOWLEDGEMENT

My first thanks go to Allah Subhanu Watala, who created us and gives us the power and the ability to learn and to conduct this research work.

My special thanks to Asst. Prof. Dr. Norsinnira Zainul Azlan and Assoc. Prof.

Dr. Salmiah Ahmad for their continuous support, encouragement and leadership, and for that, I will be forever grateful.

Also I wish to express my appreciation and thanks to those who provided their time, effort and support for this project. To the members of my dissertation committee:

thank you for sticking with me.

Finally, it is my utmost pleasure to dedicate this work to my dear parents, wife and my family, who granted me the gift of their unwavering belief in my ability to accomplish this goal: thank you for your support and patience.

viii

TABLE OF CONTENTS

Abstract ... ii

Abstract in Arabic ... iii

Approval Page ... iv

Declaration ... v

Copyright Page ... vi

Acknowledgement ... vii

Table of Contents ... viii

List of Tables ... xi

List of Figures ... xii

List of Abbreviation ... xvii

List of Symbols ... xviii

CHAPTER ONE: INTRODUCTION ... 1

1.1 Background ... 1

1.2 Problem Statement and its Significance ... 2

1.3 Research Objectives... 3

1.4 Research Methodology ... 3

1.5 Research Scope ... 4

1.6 Research Contribution ... 4

1.7 Dissertation outline ... 5

CHAPTER TWO: LITERATURE REVIEW ... 7

2.1 Introduction... 7

2.2 Anthropomorphic Hand ... 8

2.2.1 Wire-driven vs. Direct-driven Joint ... 8

2.3 Pneumatic Artificial Muscle (PAM) Actuators ... 11

2.3.1 Modelling of PAM Actuators ... 13

2.4 Control of PAM Actuators... 16

2.5 Adaptive Backstepping Control Design ... 22

2.6 Grasping Weight-varying Objects ... 27

2.6.1 Slippage Detection Technique ... 29

2.7 Summary ... 32

CHAPTER THREE: PNEUMATIC ANTHROPOMORPHIC HAND DEVELOPMENT AND MODELLING AND ADAPTIVE BACKSTEPPING CONTROL DESIGN ... 34

3.1 Introduction... 34

3.2 Mechanical Design of Anthropomorphic Robotic Hand ... 34

3.2.1 Working principle ... 36

3.2.2 Initial Anthropomorphic Prototype ... 38

3.2.3 Final Anthropomorphic Hand Prototype... 43

3.2.4 HandAssembly ... 51

3.2.5 Mechanical Properties ... 53

3.3 Mathematical Modelling of PAM Actuator and Robotic Finger ... 55

3.3.1 Static Force Model of PAM Actuator ... 55

ix

3.3.2 System Modelling for Position Controller Design ... 57

3.4 Controller Design... 58

3.4.1 Adaptive Backstepping Position Controller ... 59

3.4.2 PID Pressure Controller ... 63

3.5 Control Design for Grasping Objects with Changing Weight ... 64

3.5.1 System Model for Grasping Weight-Varying Objects ... 64

3.5.2 Control Design for Grasping Weight-Varying Objects ... 66

3.6 Summary ... 72

CHAPTER FOUR: RESULTS AND DISCUSSION ... 75

4.1 Introduction... 75

4.2 Experiment Setup... 75

4.2.1 Supply Unit ... 77

4.2.1.1 Power Sources ... 82

4.2.1.2 Pressurized Air Source ... 82

4.2.2 Control Unit ... 83

4.2.2.1 SMC ITV0031 Electro Pneumatic Regulator ... 84

4.2.2.2 NI BNC-2110 Block ... 87

4.2.2.3 National Instruments PCI-6024E Board ... 89

4.2.2.4 Target PC and Simulink Real Time Software ... 91

4.2.2.5 Host PC and Simulink Real Time Explore ... 93

4.2.3 Hand Unit ... 95

4.2.3.1 Anthropomorphic Hand ... 96

4.2.3.2 AM-01 Model PAM Actuator ... 96

4.2.3.3 404 Model Linear Position Sensor ... 98

4.2.3.4 BWY8-30 Model Extension Spring ... 100

4.2.3.5 FlexiForce A201 Model FSR Sensor... 100

4.3 Position Control ... 102

4.3.1 Simulink Model for Matlab Simulation ... 105

4.3.2 Simulink Models for Real Time Applications and Hardware Experiment... 107

4.3.3 Tuning of Control Parameters ... 108

4.3.4 Translational Motion of The PAM Actuator Linked to The Robotic Finger ... 112

4.3.5 Train Trajectory for Rotational Motion of the Robotic Fingers ... 117

4.3.6 Angle Staircase Trajectory for Rotational Motion of the Robotic Fingers... 123

4.3.7 Simultaneous Motion of the Robotic Fingers ... 130

4.4 Grasping Weight-Varying Objects ... 134

4.4.1 Grasping a 100 g Object Increased to 300 g ... 135

4.4.2 Grasping a 300 g Object Increased to 500 g ... 140

4.5 Summary ... 143

CHAPTER FIVE: CONCLUSION AND RECOMMENDATION ... 144

5.1 Conclusion ... 144

5.2 Recommendation ... 145

REFERENCES ... 146

LIST OF PUBLICATIONS ... 149

x

APPENDIX A (i): MIDDLE PULLEY DRAWING ... 150

APPENDIX A (ii): PROXIMAL PULLEY DRAWING ... 151

APPENDIX B (i): ITV0000 ELECTRO PNEUMATIC REGULATOR SPECIFICATIONS ... 152

APPENDIX B (ii): ITV0000 ELECTRO PNEUMATIC REGULATOR DIMENSIONS ... 154

APPENDIX B (iii): ITV0000 ELECTRO PNEUMATIC REGULATOR WIRING ... 155

APPENDIX C: BNC-2110 SPECIFICATIONS AND SIGNAL DESCRIPTIONS ... 157

APPENDIX D (i): NATIONAL INSTRUMENTS PCI-6024E ANALOG INPUT (D/A) ... 159

APPENDIX D (ii): NATIONAL INSTRUMENTS PCI-6024E ANALOG OUTPUT (A/D) ... 161

APPENDIX E: 404 MODEL POTENTIOMETER SPECIFICATIONS ... 162

APPENDIX F: BWY8-30 EXTENSION SPRING SPECIFICATIONS ... 164

APPENDIX G: A201 MODEL FSR SENSOR PROPERTIES ... 165

APPENDIX H: ELECTRICAL CHARACTERISTIC OF MCP602 OP- AMP ... 166

xi

LIST OF TABLES

Table No. Page No.

2.1 Comparison between direct-driven & wire-driven joint 9 3.1 Finger parts presenting schematic links and pulleys 51

3.2 Manufacturing material of hand parts 52

3.3 Mechanical properties of the robotic hand 55

4.1 Description of the electrical wire junctions 80

4.2 Signal source configuration 90

4.3 AM-01 air muscle specifications 98

4.4 Calibration parametres 104

4.5 Control and system model parameters 113

4.6 Settling time and RMSE for the train trajectory of robotic fingers 120 4.7 Settling time and RMSE for angle staircase trajectory of thumb finger 125 4.8 Settling time and RMSE for the angle for the staircase trajectory of

index and middle fingers 125

4.9 Actual PAM pressure and angular position for the staircase trajectory of

thumb finger 126

4.10 Actual PAM pressure and angular position for the staircase trajectory

of index and middle fingers 126

4.11 Comparison of simulation and experiment results for position control 130

4.12 Grasping of cylindrical shaped objects 131

xii

LIST OF FIGURES

Figure No. Page No.

1.1 Research methodology 6

2.1 Prototype of anthropomorphic hand 10

2.2 FRH-4 hand 10

2.3 Pneumatic hand with direct-driven joints actuated by 13 PAM actuators 11 2.4 Working principle of PAM actuator, (a) initial condition, (b) pressurized

condition 12

2.5 The Schematic constitution of PAM 13

2.6 The linear relationship between PAM pressure and contraction force 14 2.7 Nonlinear relationship between PAM contraction ratio and force 15 2.8 Positioning tests with a PAM actuator using a PI controller 16

2.9 Cascade system for force control 17

2.10 PID with gap system for joint angle control 17

2.11 Force response to chirp signal 18

2.12 Angle response of staircase signal 18

2.13 Experiment setup for PI controller 19

2.14 Radial angle response 19

2.15 Palmar angle response 20

2.16 Block diagram of control system with single on/off valve configuration 20

2.17 Sinusoidal tracking at 0.5 Hz 21

2.18 A comparison of three types of controller for PMA actuators 22

2.19 Grasping various shapes of objects 27

2.20 Example of the five-fingered robot hand holding an object 28 2.21 Time-pressure relationship for the robot finger holding an object (15 g) 28

2.22 Action of the robot fingers holding objects 29

xiii

2.23 Carbon Micro Coil (CMC) 30

2.24 Outputs of CMC sensor and lazer displacement meter 31

2.25 Force sensing resistors (FSR) sensor 32

3.1 Flowchart of design and fabrication process 35

3.2 Schematic diagram of the robtic finger 37

3.3 The initial mechanical drawing of the finger 38

3.4 Test of the initial finger prototype 39

3.5 The location of pulley-fixing point 40

3.6 Initial Finger prototype imitates human finger in bending 40 3.7 Mechanical drawing of the initial hand prototype, (a) hidden lines are

visible, (b) shaded with edges 41

3.8 Performing grasping task, (a) animation test in Solidworks, (b) real test 42 3.9 Anthropomorphic hand prototype (a) mechanical drawing, (b) fabricated

hand 43

3.10 Mechanical dimensions of robotic finger, (a) top view, (b) side view 44

3.11 Mechanical dimensions of the palm, (a) top view, (b) front view 45

3.12 Human being hand 46

3.13 The angles between robotic fingers, (a) middle, index and thumb, (b)

the basic angle of thumb 47

3.14 The parts of the robotic fingers, (a) middle and index, (b) thumb 48 3.15 The robotic palm, (a) tinny paths of tendons and the bases of the

fingers, (b) passing the thumb tendons (c) palm shaded with edges 49

3.16 The parts of the robotic hand 50

3.17 Designed pulley-fixing point, (a) back view, (b) real pulley 51

3.18 Assembly process of the robotic hand 54

3.19 The cascade controller structure 65

3.20 The cascade controller structure with slippage detector for force control 73

3.21 The criteria of slippage detection 74

xiv

4.1 The hardware experiment setup 76

4.2 Schematic block diagram of the hardware experiment setup 79

4.3 Labelled electrical wire junctions 81

4.4 The components of supply unit 82

4.5 VIAIR on board air system part No. 1000 83

4.6 (a) Plumbing diagram. (b) Wiring diagram 84

4.7 Control unit 85

4.8 Electro pneumatic regulator, (a) block diagram, (b) working principle 86

4.9 Connecting the BNC-2110 to the DAQ device 87

4.10 BNC-2110 front panel 88

4.11 Voltage measurement between the ground and the building earth 89

4.12 National Instruments PCI-6024E block diagram 91

4.13 Configuration of host to target communication 92

4.14 Target settings 92

4.15 Boot configuration of target PC 93

4.16 Configuration Parameters for the Simulink Real Time 94 4.17 Execute target application using Simulink Real time Explorer 95 4.18 Signal Monitoring and parameter tuning using insterument Panel 95

4.19 Activation of signal logging 96

4.20 Hand unit components 97

4.21 PAM actuator, (a) before installing on the epriment, (b) initila

condition, (c) pressurized condition 99

4.22 Wiring diagram of the ptentiometer 100

4.23 FlexiForce A201 FSR Sensor 101

4.24 Measured force at the tip of the thumb 102

4.25 Adapting circuit for the force sensor, (a) schematic diagram, (b) the buit circuit of FSR sensor 1, (c) the built circuit of FSR sensor 2, (d)

pinout of MCP 602 103

xv

4.26 FSR sensor calibration, (a) resulted graph, (b) experiment setup 104

4.27 Simulink model for Matlab simulation 106

4.28 Simulink model of PAM actuator with robotic finger 107

4.29 Simulink model for real time application 109

4.30 Simulink model of the entire hand for real time application 110

4.31 Tuning procedure for control parameters 112

4.32 Non-ideal case of adaptive backstepping position controller, (a)

c1=c2=250, (b) c1=c2=1750 114

4.33 Simulink Real Time Explorer panel for PAM translational motion 115 4.34 Position response of the PAM actautor linked to the robotic finger 117 4.35 Initial fingers angles, (a) index & middle fingers, (b) thumb finger 118 4.36 Simulink Real Time Explorer panel for the rotational motion 119 4.37 Angle train trajectory for index & middle fingers, (a) angular position

response, (b) PAM pressure 121

4.38 Angle train trajectory for thumb finger, (a) angular position response,

(b) PAM pressure 122

4.39 Mechanical response of the thumb finger for train trajectory 123 4.40 Angle staircase trajectory for index & middle fingers, (a) angular

position response, (b) PAM pressure 127

4.41 Angle staircase trajectory for index & middle fingers, (a) angular

position response, (b) PAM pressure 128

4.42 Mechanical response of index & middle fingers for staircase trajectory 129 4.43 Simulink Real Time Explorer panel for the entire hand 131 4.44 Simultaneous motion of robotic fingers, (a) angular position response,

(b) PAM pressure 132

4.45 Robotic hand performing grasping task 133

4.46 Simulink model of the entire hand including FSR sensors for real time

application 134

4.47 Simulink model of index & middle subsystem including slip detector

for real time application 136

xvi

4.48 Simulink Model of the Slippage detector 137

4.49 Simulink Real Time Explorer panel for force control 137 4.50 Grasping 100 g-300 g object, (a) force response, (b) position response,

(c) PAM pressure 138

4.51 Grasping of 100 g-300 g object, (a) before slippage detection, (b) after

slippage detection 139

4.52 Grasping 300 g - 500 g object, (a) force response, (b) position

response, (c) PAM pressure 141

4.53 Grasping of 300 g - 500 g object, (a) before slippage detection, (b) after

slippage detection 142

xvii

LIST OF ABBREVIATION

ABS Acrylonitrile Butadiene Styrene

Avg. Average

C Constant

CMC Carbon Micro‐Coil

CNC Computer Numerical Control

Deg Degree

DOF Degrees Of Freedom FSR Force Sensing Resistor LC Inductor and Capacitor

NC Not Connected

No. Number

PAM Pneumatic Artificial Muscle

PID Proportional, Integral and Derivative control

QTY Quantity

R Resistor

RMSE Root Mean Square Error

sec Second

vs. Versus

xviii

LIST OF SYMBOLS

a defined parameter for control design (Kg^-1)

a0 PAM model parameter (N)

a1 PAM model parameter (N/m)

a2 PAM model parameter (N/m²)

a3 PAM model parameter (N/m³)

b0 PAM model parameter (N/Pa)

b1 PAM model parameter (N.m^-1/Pa)

Beff an effective viscous friction factor (Ns/m) C1 & C2 adaptive backstepping control parameters (s^-1) Dobj grasped object diameter (mm)

Fc coulomb friction (N)

Fg grasp force (N)

Finitial initial grasp force (N)

Fslip slip detector output (N)

I total moment of inertia of the finger (Kg.m²) K slip detector coefficient

k1 spring coefficient (N/m) Kp, Ki & Kd PID control parameters Pm actual muscle pressure (MPa) Pmax maximum muscle pressure (MPa)

R pulley radius (m)

S slip detector sensitivity

xix Ts fundamental sample time ts settling time (s)

x1 translational position of PAM actuator

1 angular position of the robotic finger

 general unknown constant parameter

1

CHAPTER ONE INTRODUCTION

1.1 BACKGROUND

Nowadays, the development of robotic hands in terms of control and mechanical design has been the concern of many researchers. These two components allow the robotic hands to perform numerous tasks of grasping and manipulating. According to the anthropomorphic inspiration, the hand has to match the human hand in terms of performance and versatility which requires specific mechanical characteristics such as the lightweight, high power to weight ratio and compact size. Thus, a suitable mechanism for the fingers has to be used to achieve these characteristics. Moreover, the actuators that drive the hand has to be safe and have a high power to weight ratio. PAM actuators are widely used as fluidic pneumatic actuators in robotic and industrial fields due to their characteristics in terms of design and implementation, performance and anthropomorphism. On the contrary, it is difficult to obtain an accurate dynamic model for PAM actuators due to the highly nonlinearity and the inherent hysteresis characteristics.

Since most prototypes of the anthropomorphic robotic hand suffer from overdesigning, high costs and lack of robustness in terms of position and force control due to implementing simple control strategies, this research proposes a design of a cascade control system by integrating an adaptive backstepping position control together with PID pressure control. The adaptive backstepping controller can be developed by combining the parameter estimators with the control law. The design process starts at the known-stable system and "back out" new controllers that progressively stabilize each outer subsystem using Lyapunov function and the process

2

ends when the final external control is reached. A three-fingered anthropomorphic hand has been developed where these fingers are driven indirectly by Pneumatic Artificial Muscle (PAM) actuators. The proposed controller is capable to compensate for the system uncertainties improving the hand's performance for position control and to grasp objects with changing weights. At this point, it is supposed that the weight of grasped object is increased and the hand holds the object to prevent it from dropping using FSR sensors.

1.2 PROBLEM STATEMENT AND ITS SIGNIFICANCE The following problems can be stated for this research.

i. Control problem. The accurate model of PAM contracting force may face two main problems. First, the high nonlinearity of the PAM dynamics due to the compressibility of the air and the nonlinear relationship between the force produced by the PAM actuator and its contraction (displacement). Second, the inherent hysteresis behaviour in PAM actuators which is mainly caused by thread-on-thread dry friction producing system uncertainties and causing energy loss. Hence, this leads to a highly nonlinear control problem for position and force control.

ii. Grasping weight-varying objects problem. The development of a controller to be capable to grasp objects with changing weights is useful for robotics applications in service field where the weight of grasped object may change while performing a grasping task such as filling machines. The grasping force may differs based on the object’s weight. Therefore, an adaptive control strategy is needed to prevent the object from dropping.

3 1.3 RESEARCH OBJECTIVES

The objectives of this research are as follows:

i. To develop a 3-fingered anthropomorphic hand actuated by PAM actuators.

ii. To formulate an adaptive backstepping position control strategy for the anthropomorphic hand.

iii. To formulate an adaptive backstepping force control strategy for the robotic hand to grasp weight-varying objects.

iv. To validate the effectiveness of the proposed controllers through simulation and hardware experiment.

1.4 RESEARCH METHODOLOGY

The aim of designing a three-fingered anthropomorphic robotic hand is to perform a grasping task of objects with different weights and sizes in the field of service robots.

This design can be illustrated by mechanical drawing using Solidworks software. The fingers are driven by tendons and actuated pneumatically by PAM actuators. To reduce the number of actuators and lower the fabrication cost, the mechanical design can be created in a way to allow only one actuator to drive the robotic finger. The finger bends under the contraction force when PAM actuator is activated and returns to its rest position under a spring restoring force when the PAM actuator is deactivated.

This research is started with a deep literature review of anthropomorphic fingers, modelling and control of PAM actuators and grasping weight-varying objects. By developing an anthropomorphic robotic hand, the first objective will be achieved. After solving control problem, a nonlinear Adaptive Back stepping control can be designed and validated by a simulation test. If this succeeds, the second objective will be

4

achieved. When the problem of weight-varying objects is solved, the controller can be formulated to achieve third objective. Finally, an experiment test will be conducted to validate overall system and if this successes, the fourth objective will be achieved and the research process ends. A flow chart of research methodology is illustrated in Figure 1.1.

1.5 RESEARCH SCOPE

The scope of this research is considered as follows.

i. The robotic hand is limited to three fingers driven by two PAM actuators.

ii. The object and finger phalanxes are rigid bodies.

iii. It is not required to measure the total grasp force in which to deal with increasing of object's weight, only the detection of slippage signal from FSR sensor is required.

iv. One FSR sensor per subsystem is used to detect the slippage signal.

v. The diameter of the grasped object is priorly known.

1.6 RESEARCH CONTRIBUTION

This research contributes in the design of a nonlinear adaptive backstepping position controller integrated with PID pressure controller for three-fingered anthropomorphic hand driven by PAM actuators. The model of PAM actuator and robotic finger is represented as a nonlinear second order system. The proposed controller is capable to compensate for the system uncertainties to improve the hand's performance for position control and to grasp objects with changing weights.

5 1.7 DISSERTATION OUTLINE

This dissertation is organized into five chapters as follows.

The literature review is presented in Chapter Two where the anthropomorphic hand, modelling and control of PAM actuators, adaptive backstepping strategy and grasping objects with changing weights are reviewed. Chapter Three elaborates the development process of the anthropomorphic hand and modelling of the contraction force of PAM actuator. In addition, the adaptive backstepping position controller and PID pressure controller are formulated for position control and extended for grasping objects with changing weight. Chapter Four evaluates the simulation and the experimental results of the cascade controller for position and force control and describes the experimental setup which is created in the Intelligent Systems Laboratory. Finally, the conclusion and recommendation are drawn in Chapter Five.