Compiling Fuzzy Logic Control Rules to Hardware Implementations

TL;DR

This paper presents a programming environment that translates fuzzy logic control rules into hardware implementations, enabling real-time approximate reasoning for complex system control.

Contribution

It introduces methods for directly translating fuzzy control rules into hardware, including specialized chips and memory-based solutions, enhancing practical real-time fuzzy control.

Findings

Hardware implementations enable faster fuzzy inference.

Two hardware realization methods demonstrated: specialized chip and memory-based.

Facilitates practical application of fuzzy control in real-time systems.

Abstract

A major aspect of human reasoning involves the use of approximations. Particularly in situations where the decision-making process is under stringent time constraints, decisions are based largely on approximate, qualitative assessments of the situations. Our work is concerned with the application of approximate reasoning to real-time control. Because of the stringent processing speed requirements in such applications, hardware implementations of fuzzy logic inferencing are being pursued. We describe a programming environment for translating fuzzy control rules into hardware realizations. Two methods of hardware realizations are possible. The First is based on a special purpose chip for fuzzy inferencing. The second is based on a simple memory chip. The ability to directly translate a set of decision rules into hardware implementations is expected to make fuzzy control an increasingly practical approach to the control of complex systems.