Analysis and design for a class of complex control systems Part I: Fuzzy modelling and identification

S.G. Cao,N.W. Rees,G. Feng
DOI: https://doi.org/10.1016/s0005-1098(97)00010-1
IF: 6.4
1997-06-01
Automatica
Abstract:This paper is the first of two dealing with the analysis and design of a class of complex control systems. This part deals with fuzzy modelling and identification. The second paper is concerned with fuzzy controller design. We believe that these two papers show a way to analyze and design complex control systems using a combination of fuzzy logic and modern control theory. The complex systems studied in these two papers can be represented by a fuzzy aggregation of a set of local linear models. The identification can be divided into two procedures. The first procedure is the identification of the membership functions, including the determination of the number of fuzzy rules, and the estimation of the parameters in the membership functions. The second is the identification of the local linear models, including the determination of the structure and the parameters of the local rule maps and the global rule interpolation. Two examples are used to demonstrate the application of the modelling method.
automation & control systems,engineering, electrical & electronic
What problem does this paper attempt to address?