A systematic approach for designing multistage fuzzy control systems

This paper proposes a systematic approach for designing a multistage fuzzy logic controller (MFLC) for large scale nonlinear systems. In designing such a controller, the major tasks are to derive fuzzy rule bases, determine membership functions of input/output variables, and design input/output scaling factors. In this work, the fuzzy rule bases are generated by rule-generated functions which are based on the negative gradient of a system performance index. The membership functions of isosceles triangle of input/output variables are fixed in the same cardinality, and only the input/output scaling factors are optimized from a genetic algorithm based on a fitness function. As a result, the search space of the parameters is narrowed down to a small space so that the MFLC can be quickly constructed and the fuzzy rules and scaling factors can easily be determined. The performance of the proposed approach is examined by computer simulations on an inverted pendulum system. The performance of single stage structure, binary tree structure and skew-binary tree structure are compared. The binary tree structure has better performance and use fewer fuzzy rules in the illustrative example.