A Model of Wavelet Chaotic Neural Network with Applications in Optimization

Abstract

Chaotic neural networks have been applied to solve function optimization problems successfully. To improve the optimization capacity of the chaotic neural network, a new chaotic neural network model called wavelet chaotic neural network was presented by transferring sigmoid function to wavelet function. The reversed bifurcation figures of signal neural unit were given and the parameters of the new model were discussed. The wavelet function is a non-monotonic function, so the new model can spend the less time than the common chaotic neural network model in function optimization. The simulation result shows that the new chaotic neural network model is superior to the common neural network model.