A Method of Optimizing the Rule Base in the Sugeno Fuzzy Inference System Using Fuzzy Cluster Analysis

This paper develops a method of optimizing the rule base in the Sugeno-type fuzzy inference system using fuzzy cluster analysis for the design problems of control systems for complex technical objects. A hybrid control model with sequential interactions of classical and fuzzy controllers is used to validate the results. In the developed adaptive system of neuro-fuzzy inference, the rule base for the fuzzy controller is automatically formed based on the knowledge about the object obtained when it has been controlled by the classical controller, which excludes an expert's participation in the formation of the rule base. The optimization problem is solved using the method of grouping the values of the input and output signals in order to reduce the number of rules and to increase the speed of the control system of the technical object. The FCM clustering algorithm is used to group the values. As a result of the algorithm operation, not all of the set input actions arrive at the fuzzy controller input. Only cluster centers determined by fuzzy sets arrive, and the boundaries between clusters are also fuzzy. The effectiveness of the proposed grouping method for ensuring the effective control of a complex technical object under conditions of uncertainty is proved.