Fuzzy control and systems

Author: Chief Editor: Ye Tao
Publisher:
Publish Date: 1998-03-01
Features: Abstract Fuzzy control technology is one of the high-tech fields in modern industry and new product development, receiving widespread attention both domestically and internationally. This book systematically presents the principles, methods, and design techniques of fuzzy control technology. It also includes fuzzy system fuzzy identification methods and fuzzy expert systems. As a textbook for graduate students, it can also serve as a reference book for undergraduate students and engineers engaged in scientific research. Excerpt: Chapter 1 Structure of Fuzzy Control Systems 1.1 Background of the Emergence of Fuzzy Control Systems Control technology is widely applied in various industrial and technological fields, becoming one of the important means of modern high technology. With the development of control technology, control theory and methods have also evolved. In addition to the extensive research on the application prospects of PID-type controllers, significant achievements have been made in the theoretical research and application of state-space methods, stochastic methods, optimal control, filter methods, and state estimation methods. A notable feature of classical control technology is its highly precise model structure. It derives complex model equations based on the physical, chemical, and mechanical characteristics of control systems. Solving these equations requires relatively complex algorithms. Due to advancements in numerical computation and computer technology, algorithmic complexity no longer significantly affects the accuracy of practical control. However, these model equations contain numerous parameters that need to be estimated, yet there is insufficient information and information features to solve these parameters effectively. If the parameter estimation problem is not properly addressed, even the best model equations may result in a suboptimal control system. Current research on control systems often involves multivariable, nonlinear, time-varying large-scale systems, where the complexity of the systems conflicts with the precision of control technology. As L.A. Zadeh pointed out, as systems become increasingly complex, people's ability to provide precise and meaningful descriptions of them correspondingly decreases, to the point where precision and meaningfulness become almost mutually exclusive characteristics. It is practically impossible to precisely and accurately describe complex phenomena and any real physical state of systems. This forces people to seek a balance and compromise between the precision and meaningfulness of control systems and practical control systems. For example, in a car driving system, an experienced driver can freely control the car through narrow channels and avoid various obstacles, but it is quite difficult and unrealistic to establish model equations using classical control theory. Similarly, in an automatic washing machine system, the washing, cleaning, and drying processes and process selection are predefined, and people set the steps of each process based on the condition of the clothes being washed. Once set, the washing machine controls the entire washing process according to the predefined parameters. As for air conditioners, they cannot understand human sensations. When the temperature rises or falls rapidly, they cannot quickly bring the room temperature to a normal level. They always operate according to certain predefined parameters. To accelerate the heating process, some parameters must be adjusted. In the systems mentioned above, humans are understood as successful nonlinear controllers, with their control variables being time-varying functions. Human experience plays a crucial role in the success of the control process, inspiring in-depth research into control principles. This principle is based on control schemes that can incorporate human thinking, reflecting the knowledge of control processes based on human experience, and the objectives that can be achieved through some form of expression, which is also easily implementable. Such control systems avoid the precise, repetitive, and error-prone model-building process and the process of precisely estimating various parameters in the model equations. In multivariable, nonlinear, time-varying large-scale systems, people can adopt simple and flexible control methods, which is the background for the emergence of fuzzy controllers. The most important feature of fuzzy controllers is their ability to reflect human experience and common-sense reasoning rules, which are expressed in language, such as "the temperature is too high, and the rate of temperature rise is also very fast, so a large-scale cooling control strategy should be adopted." For this kind of experience expressed in language, a descriptive method must be provided, and this experience is diverse. For example, there could be an experience rule like "the temperature is slightly low, and the rate of temperature rise is very fast, so a slightly cooling control strategy should be adopted." Comprehensive consideration of numerous control strategies, i.e., common-sense reasoning rules.