Simulation Research of Fuzzy Controller with Sigmoid Scaling Factor

Abstract

For the fuzzy controller performance optimization, several optimization strategies on the membership function, control rules and scaling factors respectively are analyzed and compared, on the basis, the sigmoid function is introduced to replace the linear factor so to optimize the conventional fuzzy controller. The theoretical analysis is carried from mapping relations of sigmoid scaling factor and relations between it and the other parameters. It is shown that sigmoid scaling factor can get the same control effect to adjusting control rules or membership function. It can ensure the dynamic performance and at the same time effectively reduce the steady-state error, but the adjusting parameter is far less than the other strategies, it is easy and feasible.