Hardware implementation of fuzzy controllers-approach and constraints

Abstract

The design approaches & constraints for hardware implementation of fuzzy logic controllers are discussed here. The digital, analog & mixed analog domains are compared. Apart from these the Memory based approach & MFC based approach are discussed. Silicon area & operational speed are the major constraints for fuzzy controllers. Most hardware implementations of fuzzy controllers use piece-wise linear functions to represent membership functions. The existing analog approach for the design of fuzzy controllers offer parallel computing with lower hardware resources but they implement much lesser rules than the fully digital ones. Compared to its analog counterpart, the digital approach has greater flexibility, easier design automation, and good compatibility with other digital systems but they are more complex and need larger chip area. Analog approach can be faster & enough flexible to compete digital fuzzy approaches. Working with digital signal limits the application to low-speed problems due to sequential calculation of fuzzy logic quantities. © Springer-Verlag Berlin Heidelberg 2006.