Chaotic neural networks with sigmoid function self-feedback and its applications

A Chaotic neural network model with sigmoid function self-feedback is proposed by introducing sigmoid function into self-feedback of chaotic neural network The analyses of the optimization mechanism of the networks suggests that sigmoid function self-feedback affects the original Hopfield energy function in the manner of the sum of the multiplications of sigmoid function to the state, avoiding the network being trapped into the local minima. The energy function is constructed, and the sufficient condition for the networks to reach asymptotical stability is analyzed and is used to instruct the parameter set of the networks for solving traveling salesman problem (TSP). Simulation research on functions´ optimization and TSP indicates that the proposed networks can find the optimal solution of combinatorial optimization problems.