Using intelligent control methods to simulate control behaviors of human beings have become the main trend of modern control theory and application. The establishment and development of Extenics has provided a new method for analyzing and studying intelligent control.
Extension control is a new branch of intelligent control. In the traditional control theory, the input and output are both of quantity values. The dynamic properties of controlled objects are represented by mathematical models which are used to describe the relationship between the input quantity value and the output quantity value. It’s very difficult to solve some control problems if only the relationship between quantity values is considered. They also become contradictory problems in the area of control. To tackle these problems, the object itself and its characteristics and their transformations shall be taken into account. The basic idea of extension control is to tackle the control problems from the perspective of information transformations, namely, to take control of the acceptable degree (relevancy) of input information as the basis for determining the correcting values of control output so as to have the controlled information transformed into the acceptable range. The extension control is the rule and method to tackle the contradictory problems in the control process by utilizing extension theory and extension methods.
Professor Wang Xingyu from East China University of Science and Technology primarily put forward the basic idea, frame and principle of extension control. Li Jian and Wang Xingyu put forward a design of extension controller and discussed the achieving issue of characteristic pattern dividing and association degree calculating. Professor Jin Yihui and his student Pan Dong from Tsinghua University improved the extension controlling architecture. They put forward double-decked self-learning extension controlling frame and implemented emulation research on kinds of linear, considerable time delay and nonlinear object by applying extension control. In recent years, Chinese and overseas scholars applied extension theory to sliding mode control, nonlinear system control and adaptive control. Besides, there were many scholars who applied extension control theory to the research and emulation verification in connection with different control systems. These researches got some achievements on numerical value optimization, numerical value analysis and formal modeling aspects.
Viewed from the perspective of control, extension control simulates the ability of human in summarizing, studying and tackling the problems of incompatibility. It deals with contradictory problem, studying the flexibility and creativity in converting contradictory problem into compatibility. Its mechanism is an alternation or substitution rather than implication and matching in traditional control. Extension control simulates the process of tackling contradictory problem by human applying flexibility and creativity.
Viewed from the models, extension control does not rely on the structural information of control system, and require no mathematical models given in advance. Therefore, it’s free of deficiency resulting from implied errors in the pre-established models or structures, so it can be effectively applied in the fields which are not fully understood or even totally unknown.
Viewed from the methods, extension control is a control method for integrating qualitative analysis and quantitative analysis, and based on extension logics. Its qualitative analysis tools are extension analysis and extension transformations while its quantitative analysis tools are correlation function and excellent degree evaluation which are based on extension set. This kind of combination has provided a proper model for tackling the diverse contradictory problems during the control process.
