An optimization of the fuzzy control algorithm

Abstract

Much effort has been put into the optimization of fuzzy controllers. This is usually done by implementing the fuzzy inference engine in hardware. This paper presents an optimized fuzzy controller using a software-only approach. Increased speed is achieved by placing constraints on certain parts of the design. The effect of these constraints are discussed. The result is analogous to the fast Fourier transform (FFT). Although the FFT is a restricted form of the general transform, it is optimized for speed over generality. The same relationship holds true for the optimized fuzzy controller. The purpose of this work is to optimize the fuzzy control algorithm to achieve hardware-like performance in software. This can be done by keeping only those features that are necessary and sufficient to achieve the same results. Everything else can be eliminated.