Research on optimisation and application of improved fuzzy neural network control algorithm

Abstract

Conventional neural fuzzy design techniques address only the parameter identification, and the structure of network is established according to the knowledge of experience. Therefore, a kind of self-organising fuzzy neural network FNN is proposed based on neural network and the advantages of FNN. First, the architecture of this proposed network is investigated, then the parameters of this new scheme are optimised by intelligent computing. At last, two experiments are done to test the performance of this new scheme. Through the analysis of experiment result, it can be seen that the proposed scheme is better than commonly used network.