Most of the design procedures of Takagi-Sugeno fuzzy controllers fall into one of following categories: (i) identification of consequent part, (ii) identifying antecedent part and consequent part in stages, and (iii) design of consequent part with stability. This paper presents a new method for the simultaneous evolution of both antecedent and consequent part of Takagi-Sugeno fuzzy controller rules with stability analysis using genetic algorithm. A fitness function involving absolute mean square error (AMSE) of state variables, the status of feasibility of linear matrix inequality (FLMI) and the number of stable subsystems is used. An application example of stabilizing an inverted pendulum system is used to show the design and stability analysis. The simulation results show the performance of the proposed design method for the design of fuzzy controller for the inverted pendulum example.

fuzzy model, fuzzy controller, genetic algorithm, stability analysis, linear matrix inequality.