- ホーム
- > 洋書
- > 英文書
- > Science / Mathematics
Full Description
Comprehensive guide to streamlining the design optimization process of intelligent robots using MODENA (MOdular DEsigN Automation)
Modular Design Automation of Intelligent Robot Systems introduces MODENA (MOdular DEsigN Automation) as a new approach to the design of intelligent robots by harnessing computational intelligence techniques like genetic programming and constrained multi-objective evolutionary algorithms. It also covers different aspects of robot design, including physical structure, control, and vision systems. Case studies are included throughout the text to aid in the practical application of concepts.
The book also provides information on:
Methods in the design automation of electronic systems, micro-electro-mechanical systems (MEMS), complex mechatronic systems, and intelligent robotic systems
MODENA's advantage in avoiding the frequent trial-and-error processes inherent in traditional design methods, instead promoting the automatic discovery of innovative designs
MODENA as a more interpretable alternative to supplement ChatGPT and other generative AI based on LLMs
Automated design of intelligent robotic systems like teaching manipulators, mechanical printers, and quarter-car suspensions as well as swarm robotic systems
Modular Design Automation of Intelligent Robot Systems is an excellent reference on the subject for professionals, researchers, and students in robotics, automation, and artificial intelligence.
Contents
Acronyms ix
1 Introduction to Design Automation and Intelligent Robotic Systems 1
1.1 Background of Design Automation 1
1.1.1 Historical Evolution 1
1.1.2 Importance in Modern Engineering 3
1.2 Intelligent Robotic Systems 3
1.2.1 Definition and Components 3
1.2.2 Challenges in Design and Optimization 4
1.3 Overview of MODENA 5
1.4 Objectives and Significance of the Book 6
2 Foundations of MODENA 9
2.1 Principles of Modular Design Automation 9
2.2 Evolutionary Computation in MODENA 11
2.2.1 Fundamentals 11
2.2.2 Application in Design Automation 18
2.3 Neural Architecture Search 29
2.3.1 Theory and Algorithms 30
2.3.2 Application in Design Automation 31
2.4 Causal Discovery 35
2.5 Comparison with Traditional and LLM-based Methods 38
2.5.1 Contrast with Traditional Approaches 38
2.5.2 Differences from LLM-based Methods 39
3 Optimization Methods in MODENA 41
3.1 General Framework of PPS 42
3.2 PPS-MOEA/D 46
3.2.1 Basic Description of PPS-MOEA/D 46
3.2.2 Experimental Study 53
3.2.3 Conclusion 63
3.3 PPS-M2M 63
3.3.1 Basic Description of M2M Strategy 64
3.3.2 PPS Search Strategy 64
3.3.3 Combination of PPS with M2M 66
3.3.4 Key Differences Between Two PPS-based Algorithms 69
3.3.5 Experiment Results 69
3.3.6 Conclusion 84
3.4 SA-PPS 85
3.4.1 Proposed Method 85
3.4.2 Push Search Stage 88
3.4.3 Pull Search Stage 89
3.4.4 BatchBALDS 95
3.4.5 Experimental Study 98
3.4.6 Conclusion 104
3.5 Genetic U-Net 105
3.5.1 Search Space and Encoding Mechanism 106
3.5.2 Evolutionary Algorithm 110
3.5.3 Experimental Setup 117
3.5.4 Results and Analysis 119
3.5.5 Conclusion 123
3.6 Evolving Hybrid Bond Graph Using GP 123
3.6.1 Bond Graph and GP 126
3.6.2 Basic Primitives 128
3.6.3 Genetic Operators 133
3.6.4 Conclusion 137
4 Application of MODENA 139
4.1 Morphology Design Automation 139
4.1.1 Teaching Manipulator Design 140
4.1.2 Optimal Design of Drive Mechanism for Electric Typewriter 153
4.2 Controller Design Automation 164
4.2.1 Discrete Controller Design for Hybrid Mechatronic Systems 164
4.2.2 Evolution Design for Vehicle Suspension System 185
4.3 Vision System Design Automation 197
4.3.1 Optimization Design for the Retinal Vascular System 197
4.3.2 Causal Feature Selection for Strabismus Diagnosis Using GNN 210
5 MODENA in Swarm Robotics 217
5.1 Introduction 217
5.2 Automated Swarm Pattern Generation for Swarm Robots 218
5.2.1 Background and Problem Formulation 218
5.2.2 Automated GRN Model Structure Design 221
5.2.3 Experimental Results 228
5.2.4 Conclusion 240
5.3 Swarm Control in Communication-denied Environments 240
5.3.1 Method Architecture of VG-Swarm 241
5.3.2 Experimental Setup 249
5.3.3 Experimental Results 251
5.3.4 Conclusion 254
5.4 Vision-based Distributed Multi-UAV Collision Avoidance 255
5.4.1 Related Work 256
5.4.2 Proposed Method 257
5.4.3 Experiments and Results 262
5.4.4 Conclusion 268
5.5 Multi-UAV Collision Avoidance via Causal Representation Learning 268
5.5.1 Related Work 268
5.5.2 Proposed Method 271
5.5.3 Experiment Results and Analysis 274
5.5.4 Conclusion 278
6 Conclusions and Future Directions 279
6.1 Conclusions 279
6.2 Research Prospects 279
References 283
Index 319
