Optimization Algorithm of Multi-Level Discrete Fuzzy Neural Networks for Solving Global Stability

To solve the problems of low adaptability for mutative external environment existed in the theory of fuzzy sets and the disadvantage of failing to obtain eventual relationship among the fuzzy sets existed in the model of neural networks, this paper proposes a multi-level discrete fuzzy neural networks model which has the "if-then" rule of fuzzy inference system. The essential stability conditon of the model is analyzed.The model uses Hard C-Means (HCM) clustering to obtain fuzzy neural networks' eventual relationship among the fuzzy sets and applies evolutionary algorithm to optimize the model. The simulation results show the effectiveness of the proposed model.