Optimal Input Vector Based Fuzzy Controller Rules Design

The paper deals with the method of design of a fuzzy controller the rules of which are based on generating the optimal input vector using a genetic algorithm. The method is first demonstrated on a simple linear system and is then applied to the start-up of a drive with a three-phase asynchronous motor with constant torque, representing a strongly nonlinear fifth order dynamic system. The proposed controller is verified through simulation using the MATLAB software package. Achieved results present a simple applicability of this proposed procedure for a wide class of nonlinear dynamic black-box systems.