A novel chaotic neural network for function optimization

Chaotic neural networks have been proved to be powerful tools to solve the optimization problems. And the chaotic neural networks whose activation function is non-monotonous will be more effective than Chen's chaotic neural network in solving optimization problems, especially in searching global minima of continuous function and traveling salesman problems. In this paper, a novel chaotic neural network for function optimization is introduced. In contrast to the Chen's chaotic neural network, the activation function of the novel chaotic neural network is wavelet function and the different-parameters annealing function are adopted in different period, so it performs extremely better when compared to the convergence speed and the accuracy of the results. And two elaborate examples of function optimization are given to show its superiority. This chaotic neural network can be a new powerful approach to solving a class of function optimization problems.