Use of fuzzy neural networks with grey relations in fuzzy rules partition optimization

Abstract

The objective of this paper is to use the back-propagation (BP) algorithm in conjunction with grey relations to find the optimal partitions of the consequent part in fuzzy neural networks (FNN). A BP algorithm with grey relational coefficient (GRC) is proposed in order to decrease the square errors of the FNN for acquiring the optimal partitions of the consequent part of fuzzy rules. From the simulation results, we find that the present method applied for fuzzy logic control of an inverted pendulum has better performance than that of the traditional BP algorithm. © Springer-Verlag Berlin Heidelberg 2005.