A Fuzzy Relational Identification Algorithm and Its Application to Predict The Behaviour of a Motor Drive System

TL;DR

This paper introduces a new fuzzy relational identification algorithm that models system behavior using simplified max-min equations, incorporating noise attenuation and applied to motor drive system prediction.

Contribution

It presents a novel fuzzy relational algorithm with an adaptation method and noise filtering, specifically designed for system behavior prediction.

Findings

Accurately predicts motor drive system behavior.

Effectively attenuates noise in relational models.

Demonstrates good potential in real-world applications.

Abstract

Fuzzy relational identification builds a relational model describing systems behaviour by a nonlinear mapping between its variables. In this paper, we propose a new fuzzy relational algorithm based on simplified max-min relational equation. The algorithm presents an adaptation method applied to gravity-center of each fuzzy set based on error integral value between measured and predicted system output, and uses the concept of time-variant universe of discourses. The identification algorithm also includes a method to attenuate noise influence in extracted system relational model using a fuzzy filtering mechanism. The algorithm is applied to one-step forward prediction of a simulated and experimental motor drive system. The identified model has its input-output variables (stator-reference current and motor speed signal) treated as fuzzy sets, whereas the relations existing between them are described by means of a matrix R defining the relational model extracted by the algorithm. The results show the good potentialities of the algorithm in predict the behaviour of the system and attenuate through the fuzzy filtering method possible noise distortions in the relational model.