Informatics in Control, Automation and Robotics I (eBook)

TABLE OF CONTENTS 5
PREFACE 10
CONFERENCE COMMITTEE 11
INVITED SPEAKERS 15
ROBOT-HUMAN INTERACTION 16
1 INTRODUCTION 16
1.1 Human Integration 17
2 INVASIVE NEURAL INTERFACE 18
2.1 Surgical Procedure 18
2.2 Neural Stimulation and Neural Recordings 19
3 NEURAL INTERACTION WITH TECHNOLOGY 20
4 CONCLUSIONS 21
INDUSTRIAL AND REAL WORLD APPLICATIONS OF ARTIFICIAL NEURAL NETWORKS 24
1 INTRODUCTION 24
2 FROM NATURAL TO ARTIFICIAL 25
2.1 Back-Propagation Learning Rule and Multi- Layer Perceptron 27
2.2 Kernel Functions Based Neural Models 28
3 ANN BASED SOLUTIONS FOR INDUSTRIAL ENVIRONMENT 29
3.1 MLP Based Adaptive Controller 29
3.2 Kernel Functions ANN Based Image Processing for Industrial Applications 32
3.3 Bio-inspired Multiple Neural Networks Based Process Identification 36
4 CONCLUSION 38
ACKNOWLEDGEMENTS 38
REFERENCES 38
THE DIGITAL FACTORY 40
1 INTRODUCTION 40
2 DEFINITION 40
3 GLOBAL DATA BASE 41
4 WORKFLOW MANAGEMENT 41
5 SUBCONTRACTORS AND COLLABORATIVE ENDINEERING 41
6 DANGER AND UNRESOLVED PROBLEMS 42
7 SUMMARY 42
REFERENCES 42
WHAT'S REAL IN "REAL-TIME CONTROL SYSTEMS"? 44
1 INTRODUCTION 44
2 MODEL CHECKING 44
3 VISUAL FORMALISM, STATECHARTS, AND STATEMATE 44
4 TIMED AUTOMATA 45
5 REAL-TIME LOGIC, GRAPHTHEORETIC ANALYSIS, AND MODECHART 46
6 PROCESS ALGEBRA 46
7 REAL-TIME RULE-BASED DECISION SYSTEMS 47
8 REAL-TIME DECISION SYSTEMS 47
SUFFICIENT CONDITIONS FOR THE STABILIZABILITY OF MULTI-STATE UNCERTAIN SYSTEMS, UNDER INFORMATION CONSTRAINTS 50
1 INTRODUCTION 50
1.1 Main Contributions of the Paper 50
2 PROBLEM FORMULATION 51
2.1 Description of Uncertainty in the Plant 51
2.2 Statistical Description of a(k) 52
2.3 Functional Structure of the Feedback Interconnection 52
2.4 Problem Statement and M-th Moment Stability 52
2.5 Motivation for our Definition of Stochastic Link and Further Comments on the Information Pattern 52
3 SUFFICIENCY CONDITIONS FOR THE ROBUST STABILIZATION OF FIRST ORDER LINEAR SYSTEMS 54
3.1 The Deterministic Case 55
3.2 Suficient Condition for the Stochastic Case 56
4 SUFFICIENT CONDITIONS FOR A CLASS OF SYSTEMS OF ORDER HIGHER THAN ONE 57
4.1 Description of the Nominal Plant and Equivalent Representations 57
4.2 Description of Uncertainty and Robust Stability 59
4.3 Feedback Structure and Channel Usage Assumptions 59
4.4 Construction of a Stabilizing Feedback Scheme 60
4.5 Suf.ciency for the Deterministic/Time-InvariantCase 60
4.6 Suf.ciency for the StochasticCase 61
4.7 Suf.ciency for the case ¯za = 0 62
4.8 Solving the Allocation Problem for a Class of Stochastic Systems 62
ACKNOWLEDGEMENTS 63
REFERENCES 63
Part 1 Intelligent Control Systems and Optimization 65
DEVICE INTEGRATION INTO AUTOMATION SYSTEMS WITH CONFIGURABLE DEVICE HANDLER 66
1 INTRODUCTION 66
2 DEVICE HANDLER TYPES 67
2.2 Configurable Device Handler 68
2.1 Device Specific Handler 68
2.3 Related Work 68
3 CDH DEVICE INTEGRATION 69
4 OFFLINE CONFIGURATION 69
4.1 Physical Line 70
4.2 Device Variables 70
4.3 Telegrams 70
4.4 Sequences 70
4.5 Global Conditions 71
5 ON-LINE USAGE 71
6 FURTHER DEVELOPMENT 72
6.1 Calculation Capability 72
6.2 Automatic Detection of Devices 72
6.3 Multi-Line Connection 72
6.4 Binary Protocols 72
6.5 Capability Profiles 73
7 CONCLUSION 73
REFERENCES 73
NON LINEAR SPECTRAL SDP METHOD FOR BMI-CONSTRAINED PROBLEMS : APPLICATIONS TO CONTROL DESIGN 74
1 INTRODUCTION 74
2 NONLINEAR SPECTRAL SDP METHOD 75
2.1 General Outline 75
2.2 The mechanism of the Algorithm 77
3 DERIVATIVES OF SP FUNCTIONS 77
3.1 Second Derivatives 78
3.2 Multiplier Update Rule 80
3.3 Solving the Subproblem - Implementational Issues 80
4 NUMERICAL EXAMPLES 81
4.1 Static Output Feedback Controller Synthesis 81
4.2 Miscellaneous Examples 82
5 CONCLUSION 84
REFERENCES 84
A STOCHASTIC OFF LINE PLANNER OF OPTIMAL DYNAMIC MOTIONS FOR ROBOTIC MANIPULATORS 86
1 INTRODUCTION 86
2 PROBLEM STATEMENT 87
3 REFORMULATION OF THE PROBLEM 88
3.1 Scaling 88
4 STRATEGY OF RESOLUTION 90
5 NUMERICAL RESULTS 91
6 CONCLUSION 91
REFERENCES 92
ACKNOWLEDGEMENTS 92
FUZZY MODEL BASED CONTROL APPLIED TO IMAGE-BASED VISUAL SERVOING 94
1 INTRODUCTION 94
2 IMAGE-BASED VISUAL SERVOING 94
2.1 Modeling the Image-Based Visual Servoing System 95
2.2 Controlling the Image-Based Visual Servoing System 95
3 PROBLEM STATEMENT 95
4 INVERSE FUZZY MODELING 96
4.1 Fuzzy Modeling 96
4.2 Inverse Modeling 97
5 FUZZY COMPENSATION OF STEADY-STATE ERRORS 97
5.1 Derivation of Fuzzy Compensation 97
6 EXPERIMENTAL SETUP 98
6.1 Vision System 98
6.2 Robotic Manipulator System 99
6.3 Systems Integration 99
7 RESULTS 99
7.1 Inverse Fuzzy Modeling 99
7.2 Control Results 100
8 CONCLUSIONS 101
ACKNOWLEDGEMENTS 101
REFERENCES 101
AN EVOLUTIONARY APPROACH TO NONLINEAR DISCRETE-TIME OPTIMAL CONTROL WITH TERMINAL CONSTRAINTS 102
1 INTRODUCTION 102
2 OPTIMAL CONTROL OF NONLINEAR DISCRETE TIME DYNAMICAL SYSTEMS 103
3 DISCRETE VELOCITY DIRECTION PROGRAMMING FOR MAXIMUM RANGE WITH GRAVITY AND THRUST 105
4 NECESSARY CONDITIONS FOR ANOPTIMUM 107
5 AN EVOLUTIONARY APPROACH TO OPTIMAL CONTROL 108
6 CONCLUSIONS 109
REFERENCES 110
A DISTURBANCE COMPENSATION CONTROL FOR AN ACTIVE MAGNETIC BEARING SYSTEM BY A MULTIPLE FXLMS ALGORITHM 112
1 INTRODUCTION 112
2 SYSTEM MODEL 113
3 CONTROLLER DESIGN 114
4 EXPERIMENTS 116
5 CONCLUSION 117
ACKNOWLEDGEMENTS 117
REFERENCES 117
AN INTELLIGENT RECOMMENDATION SYSTEM BASED ON FUZZY LOGIC 118
1 INTRODUCTION 118
2 RECOMMENDATION SYSTEM 118
2.1 System Architecture 118
2.2 Information Description 119
2.3 Application of Fuzzy Logic Control Theory in a Recommendation System 119
3 FUZZY RECOMMENDATION 119
3.1 Fuzzy Information Database 119
3.2 Similarity Matching 120
3.3 Filtering & Ranking
3.4 Profiling Agent 120
4 EXAMPLE 121
4.1 Similarity Matching and Preference Learning Example 121
4.2 Application Scenario 121
5 CONCLUSION 121
REFERENCES 122
MODEL REFERENCE CONTROL IN INVENTORY AND SUPPLY CHAIN MANAGEMENT 124
1 INTRODUCTION 124
2 MODEL PREDICTIVE CONTROL 124
2.1 Implementing the Cost Function 125
3 SIMULATIONS 126
3.1 Simulator Implementation Tool 126
3.2 Inventory Control Simulations 127
3.3 Step Response Simulations 128
3.4 Simulations with a More Realistic Demand Pattern 128
4 CONCLUSIONS REFERENCES 129
AN LMI OPTIMIZATION APPROACH FOR GUARANTEED COST CONTROL OF SYSTEMS WITH STATE AND INPUT DELAYS 130
1 INTRODUCTION 130
2 PROBLEM STATEMENT AND DEFINITIONS 131
3 SOLUTION IN THE LMI FRAMEWORK 132
4 EXAMPLE 135
5 CONCLUSIONS 135
REFERENCES 135
USING A DISCRETE-EVENT SYSTEM FORMALISM FOR THE MULTI- AGENT CONTROL OF MANUFACTURING SYSTEMS 138
1 INTRODUCTION 138
2 THE MULTI-AGENT SYSTEMS CONTROL APPROACH 139
3 THE DISCRETE-EVENT MODELLING FRAMEWORK 139
4 THE INTERACTIONS OF A PA WITH WAS AND TAS 140
5 SOME EXPERIMENTAL RESULTS AND FUTURE PLANS 144
6 CONCLUDING REMARKS 145
REFERENCES 145
Part 2 Robotics and Automation 147
FORCE RIPPLE COMPENSATOR FOR A VECTOR CONTROLLED PM LINEAR SYNCHRONOUS MOTOR 148
1 INTRODUCTION 148
2 SIMULATION MODEL 149
2.1 Model of LSM 149
2.2 Non-idealities of PMLSM 150
2.3 Current Controller of the Linear Motor 151
2.4 Verification of the Simulation Model 151
3 DISTURBANCE COMPENSATION 153
4 CONCLUSION 154
APPENDIX 154
REFERENCES 154
HYBRID CONTROL DESIGN FOR A ROBOT MANIPULATOR IN A SHIELD TUNNELING MACHINE 156
1 INTRODUCTION 156
2 MODELING 157
2.1 Manipulator Model 157
2.2 Environment Model 158
2.3 Hydraulic Thrust Cylinders 159
3 HYBRID CONTROLLER 159
3.1 Feedback Linearization 160
3.2 Transformation of Sensor Data 160
3.3 Selection Matrix 160
3.4 Position Control 161
3.5 Force Control 161
3.6 Impedance Control 161
3.7 Simulation Results 162
4 CONCLUSIONS 162
REFERENCES 163
MOCONT LOCATION MODULE: A CONTAINER LOCATION SYSTEM BASED ON DR/ DGNSS INTEGRATION 164
1 INTRODUCTION 164
2 OVERVIEW OF THE MOCONT SYSTEM 165
4 DR SUBSYSTEM DESIGN 167
4.1 Description of the DR Subsystem 167
4.2 DR/DGNSS Integration 167
5 EXPERIMENTAL RESULTS 169
6 CONCLUSIONS 170
ACKNOWLEDGEMENTS 171
REFERENCES 171
PARTIAL VIEWS MATCHING USING A METHOD BASED ON PRINCIPAL COMPONENTS 172
1 INTRODUCTION 172
2 OVERALL DESCRIPTION OF THE METHOD 173
3 PRINCIPAL COMPONENTS DATABASE GENERATE 174
3.1 Virtual Partial Views Computation 174
3.2 Principal Components of VPV Computation 176
4 MATCHING PROCESS 176
4.1 Initial Transformation Matrix Computation 176
4.2 ICP Algorithm Application 177
5 EXPERIMENTAL RESULT 178
6 CONCLUSIONS 179
ACKNOWLEDGEMENTS 179
REFERENCES 179
TOWARDS A CONCEPTUAL FRAMEWORK-BASED ARCHITECTURE FOR UNMANNED SYSTEMS 180
1 INTRODUCTION 180
2 DESCRIPTION OF THE SOFTWARE ARCHITECTURE 181
2.1 Architectural Elements 182
2.2 Middleware-Based Framework 183
3 DESIGN OF AN AUTONOMOUS ENTITY 184
3.1 Structure of an AE 184
3.2 Provider of Concrete Services 185
3.3 Retrieving CSPs 186
3.4 Provision of Services 186
4 EXPERIMENTAL RESULTS 186
5 CONCLUSIONS 188
REFERENCES 189
A INTERPOLATION-BASED APPROACH TO MOTION GENERATION FOR HUMANOID ROBOTS 192
1 INTRODUCTION 192
2 HUMANOID ROBOT AND THE TARGET MOTIONS 192
2.1 Humanoid Robot 192
2.2 Tai Chi Chuan 192
3 MOTION GENERATION SYSTEM 193
3.1 Interpolation-based Motion Generation 193
3.2 Classification of the Postures in Balance Space 193
4 EXPERIMENT 196
4.1 Performance Results 196
4.2 Effectiveness of Balance Checker 196
5 CONCLUSION 197
ACKNOWLEDGEMENTS 197
REFERENCES 198
REALISTIC DYNAMIC SIMULATION OF AN INDUSTRIAL ROBOT WITH JOINT FRICTION 200
1 INTRODUCTION 200
2 FINITE ELEMENT REPRESENTATION OF THE MANIPULATOR 201
3 THE DRIVING SYSTEM 202
4 JOINT FRICTION MODEL 203
5 CLOSED-LOOP ROBOT MODEL 203
6 PERTURBATION METHOD 204
7 SIMULATION RESULTS 205
8 CONCLUSIONS 206
ACKNOWLEDGEMENTS 207
REFERENCES 207
A NEW PARADIGM FOR SHIP HULL INSPECTION USING A HOLONOMIC HOVER- CAPABLE AUV 208
1 INTRODUCTION AND EXISTING CAPABILITIES 208
2 PHYSICAL VEHICLE OVERVIEW 209
3 OUR APPROACH TO HULL NAVIGATION 209
3.1 Suitability of the DVL for this Task 210
3.2 Two Approaches Using “Slicing” 211
3.3 Role of Low- and Mid-Level Control 211
4 SUMMARY 212
REFERENCES 212
DIMSART: A REAL TIME - DEVICE INDEPENDENT MODULAR SOFTWARE ARCHITECTURE FOR ROBOTIC AND TELEROBOTIC APPLICATIONS 214
1 INTRODUCTION 214
1.1 Telepresence Environment 214
1.2 Control and Software engineering Interplay 215
1.3 Existing Architectures 215
2 ROBOT CONTROL 215
2.1 Wave Variables Scheme as Bilateral Control Example 215
2.2 General Robot Control setup 216
2.3 Bilateral Control Scheme with DIMSART Embedded 217
3 ARCHITECTURE OVERVIEW 217
3.1 Modules 217
3.2 The Data Base 218
3.3 The Module Engine 219
3.4 The Frame 219
4 EXAMPLE 220
5 CONCLUDING REMARKS AND REMAINING ASPECTS 220
REFERENCES 221
Part 3 Signal Processing, Systems Modeling and Control 223
ON MODELING AND CONTROL OF DISCRETE TIMED EVENT GRAPHS WITH MULTIPLIERS USING (MIN, +) ALGEBRA 224
1 INTRODUCTION 224
2 RECURRENT EQUATIONS OF TEGM’s 225
3 DIOID, OPERATORIAL REPRESENTATION 225
4 JUST IN TIME CONTROL 227
4.1 Residuation Theory 227
4.2 Control Problem Statement 228
5 CONCLUSION 228
REFERENCES 229
MODEL PREDICTIVE CONTROL FOR HYBRID SYSTEMS UNDER A STATE PARTITION BASED MLD APPROACH ( SPMLD) 230
1 INTRODUCTION 230
2 HYBRID SYSTEMS MODELING 231
2.1 Mixed Logical Dynamical Model 231
2.2 Piecewise Affine Model 231
3 MODEL PREDICTIVE CONTROL 231
3.1 Model Predictive Control for the MLD Systems 232
3.2 Model Predictive Control for the PWA Systems 232
4 MPC FOR STATE PARTITION BASED MLD ( SPMLD) FORMALISM 233
4.1 The SPMLD Formalism 233
4.2 Reformulation of the MPC Solution 233
4.3 Compared Computational Burden 234
4.4 Further Improvements of the Optimization Time 235
5 APPLICATION 235
5.1 Description of the Benchmark 235
5.2 Application of MPC for the SPMLD Formalism 235
6 CONCLUSION 237
REFERENCES 237
EFFICIENT SYSTEM IDENTIFICATION FOR MODEL PREDICTIVE CONTROL WITH THE ISIAC SOFTWARE 238
1 INTRODUCTION 238
2 SYSTEM IDENTIFICATION AND MODEL PREDICTIVE CONTROL 238
3 ISIAC 240
3.1 Approaches to Model Estimation 240
3.2 General Structure and Layout 241
4 WORKING WITH ISIAC 242
4.1 Experiment Design 242
4.2 Working with Data 242
4.3 Working with Models 242
4.4 Estimating and Validating Models 242
4.5 Building the Control Model 243
5 AN INDUSTRIAL APPLICATION: MODEL PREDICTIVE CONTROL OF A MTBE UNIT 243
6 CONCLUSION 244
REFERENCES 245
IMPROVING PERFORMANCE OF THE DECODER FOR TWO-DIMENSIONAL BARCODE SYMBOLOGY PDF417 246
1 INTRODUCTION 246
2 LOCALIZING THE DATA REGION 246
3 DECODING CODEWORDS FROM BAR- SPACE PATTERNS 247
4 EXPERIMENTAL RESULTS 248
5 CONCLUSION 249
REFERENCES 249
CONTEXT IN ROBOTIC VISION 252
1 INTRODUCTION 252
2 CONTEXT IN COMPUTER VISION 253
3 AN OPERATIVE DEFINITION OF CONTEXT 254
3.1 Model Set M 254
3.2 Operator Set Z 254
3.3 Contextual Changes 255
4 BAYESIAN CONTEXT SWITCHING 255
4.1 Opportunistic Switching 255
4.2 Context Commutation 256
4.3 A Practical Implementation 257
5 CONCLUSIONS 259
REFERENCES 259
DYNAMIC STRUCTURE CELLULAR AUTOMATA IN A FIRE SPREADING APPLICATION 260
1 INTRODUCTION 260
2 BACKGROUND 261
3 DSCA MODELLING 262
4 DSCA SIMULATION 263
ACTIVITY TRACKING 263
6 FIRE SPREADING APPLICATION 265
7 CONCLUSION 267
REFERENCES 267
ACKNOWLEDGEMENTS 267
SPEAKER VERIFICATION SYSTEM 268
1 INTRODUCTION 268
2 BASIC IDEA OF THE VQ-VERIFICATION 269
3 THE GMM BASED SPEAKER VERIFICATION 270
4 EXPERIMENTAL RESULTS 271
5 CONCLUSION 273
REFERENCES 273
MOMENT-LINEAR STOCHASTIC SYSTEMS 276
1 INTRODUCTION 276
2 MLSS: FORMULATION AND BASIC ANALYSIS 277
3 FURTHER RESULTS 279
4 EXAMPLE: MJLS 280
5 MLSS MODELS FOR NETWORK DYNAMICS: SUMMARY 282
REFERENCES 283
ACTIVE ACOUSTIC NOISE CONTROL IN DUCTS 286
1 INTRODUCTION 286
2 FEEDFORWARD CONTROL 287
3 EXPERIMENTAL SET-UP 290
4 IDENTIFICATION 291
5 EXPERIMENTAL RESULTS 291
6 CONCLUSIONS 293
REFERENCES 293
HYBRID UML COMPONENTS FOR THE DESIGN OF COMPLEX SELF-OPTIMIZING MECHATRONIC SYSTEMS 294
1 INTRODUCTION 294
2 RELATEDWORK 295
3 MODELING RECONFIGURATION 295
4 THE APPROACH 297
4.1 Hybrid UML Model 298
4.2 Hybrid Statecharts 298
4.3 Hybrid Components 298
4.4 Modular Recon.guration 299
5 RUN-TIME ARCHITECTURE 300
6 CONCLUSION AND FUTURE WORK 300
REFERENCES 300
AUTHOR INDEX 302

ROBOT-HUMAN INTERACTION (p. 3)

Practical experiments with a cyborg

Kevin Warwick
Department of Cybernetics, University of Reading,
Whiteknights, Reading, RG6 6AY, UK
Abstract: This paper presents results to indicate the potential applications of a direct connection between the human nervous system and a computer network. Actual experimental results obtained from a human subject study are given, with emphasis placed on the direct interaction between the human nervous system and possible extra-sensory input.

An brief overview of the general state of neural implants is given, as well as a range of application areas considered. An overall view is also taken as to what may be possible with implant technology as a general purpose human-computer interface for the future.

1 INTRODUCTION

There are a number of ways in which biological signals can be recorded and subsequently acted upon to bring about the control or manipulation of an item of technology, (Penny et al., 2000, Roberts et al., 1999). Conversely it may be desired simply to monitor the signals occurring for either medical or scientific purposes.

In most cases, these signals are collected externally to the body and, whilst this is positive from the viewpoint of non-intrusion into the body with its potential medical side-effects such as infection, it does present enormous problems in deciphering and understanding the signals observed (Wolpaw et al., 1991, Kubler et al., 1999).

Noise can be a particular problem in this domain and indeed it can override all other signals, especially when compound/collective signals are all that can be recorded, as is invariably the case with external recordings which include neural signals.

A critical issue becomes that of selecting exactly which signals contain useful information and which are noise, and this is something which may not be reliably achieved. Additionally, when specific, targeted stimulation of the nervous system is required, this is not possible in a meaningful way for control purposes merely with external connections.

The main reason for this is the strength of signal required, which makes stimulation of unique or even small subpopulations of sensory receptor or motor unit channels unachievable by such a method.

A number of research groups have concentrated on animal (non-human) studies, and these have certainly provided results that contribute generally to the knowledge base in the field. Unfortunately actual human studies involving implants are relatively limited in number, although it could be said that research into wearable computers has provided some evidence of what can be done technically with bio-signals.

We have to be honest and say that projects which involve augmenting shoes and glasses with microcomputers (Thorp, 1997) are perhaps not directly useful for our studies, however monitoring indications of stress or alertness by this means can be helpful in that it can give us an idea of what might be subsequently achievable by means of an implant.

Of relevance here are though studies in which a miniature computer screen was fitted onto a standard pair of glasses.