Machine intelligence control method

Author: Wang Hao
Publisher:
Publish Date: 2002-04-01
Features: Based on an analysis of the current state of industrial robot execution-level control technology, this book elaborates on intelligent methods for robot operation control, primarily including fuzzy control, variable structure control, neural network control, and their integrated applications in the theoretical and practical aspects of trajectory tracking for multi-degree-of-freedom rigid robots. It describes a type of fuzzy variable structure control system that can effectively utilize the mathematical model of multi-degree-of-freedom rigid robots in the form of differential equations while accommodating behavioral models described in natural language. The book summarizes the similarities between fuzzy supervisory control and typical variable structure control, revealing the fuzzy nature of boundary layer methods in variable structure control that are effective in reducing chattering. Additionally, the book proposes new concepts and ideas in the context of general fuzzy control systems. For example, the maximum membership degree principle can unify various "fragmented" fuzzy mathematical operations in current fuzzy control methods, such as the acquisition of fuzzy relations, the synthesis of fuzzy relations, and the projection of fuzzy sets, into a common theoretical framework. It derives geometric selection criteria for redundant design parameters of a class of fuzzy supervisory control algorithms and defines the distance equivalence between fuzzy rules to achieve efficient reasoning processes. The book also proposes strategies and algorithms for dynamically constructing fuzzy controllers. These new strategies and algorithms have been applied in high-efficiency fuzzy variable structure controllers for robots. This book is suitable for researchers, designers, and practitioners engaged in the intelligent control systems of robots, as well as for students and faculty in relevant disciplines at universities.