A cad tool for simulation of fuzzy controllers

Abstract EN

This paper presents a user-friendly, menu-driven simulation tool for fuzzy user-defined controllers.

Fuzzy controllers as well as other fuzzy systems are normally used when adequate crisp mathematical models for the process under control are not available because of process complexity and non-linearities, and when the system variables can naturally be fuzzified.

The hardest part in the design of a fuzzy control system lies in the assignment and adjustment of input and output membership functions, the inference rules and also in the selection of a suitable defuzzification strategy.

The software package allows the user to systematically design, save the description of, simulate, test, evaluate, and improve the performance of any fuzzy controller.

This can be accomplished through an easy to implement trial-and-error methodology whereby the fuzzy model can be tuned to yield the desired output response.

The paper also compares the results of a new defuzzification strategy with the minimax weighted average formula which has been utilized as a defuzzification method in many fuzzy control situations.