VLSI FUZZY PROCESSORS

Jaime Ramirez-Angulo (1) and Angel Rodriguez Vazquez (2)

(1) Klipsh Department of Electrical and Computer Engineering
New Mexico State University, Las Cruces, New Mexico, USA
(2) Department of Analog Design
Centro Nacional de Microelectronica, Seville, Spain

This course discusses various approaches for the VLSI implementation of fuzzy 
logic controllers. Special emphasis is given to the implementation of high
speed analog fuzzy processors with fully programmable characteristics.
The architectures discussed allow modular expansion in order to cater for any
number of fuzzy inference rules and any number of input/output variables
required by the application. The basic building blocks for the fuzzy logic
processors are discussed; and several approaches for defuzzification ar 
presented. Applications of these systems are discussed.

General outline of course:

1. Introduction to fuzzy logic.
2. Membership function circuits with programmable characteristics.
3. Min-Max circuits.
4. Voltage-mode, current-mode and charge-mode defuzzification schemes.
5. Architecture of modular VLSI fuzzy rule-defuzzifier chips.
6. Other approaches of digital and analog VLSI fuzzy logic hardware.

Course duration: 6 hours (full day).