Implementing fuzzy rule-based systems on silicon chips

The authors address the implementation of a fuzzy simulator (FSIM) and discuss architectures for a general-purpose VLSI fuzzy-inference processor. The FSIM tool aids in the rapid prototyping of fuzzy production system (FPS), and represents a convenient transitional step in the implementation of an FPS on silicon. They present a brief theoretical review of fuzzy reasoning, introduce the FSIM, and discuss FPS development using the FSIM. An overall picture of the various stages involved in developing a fuzzy-inference processor is provided. The authors then outline the general architecture of a VLSI inference processor for FPSs. To further illustrate the development of a fuzzy inference processor, they describe an FPS example from conceptualization to implementation on silicon chips.<>