A novel maximum neural network with stochastic dynamics for N-queens problems

We propose a novel maximum neural network with stochastic dynamics for solving NP-hard optimization problems, the N-Queens problems. A self-feedback term with stochastic characteristic is introduced into motion function of the maximum neural network, which increases the dynamics of the neural network to search for globally optimal solutions. Moreover, several new constraints having random selection character are presented and used in the proposed algorithm to drive the network to escape from local minima. With the stochastic dynamics and those new constraints, the proposed algorithm has a great ability to find optimal or near-optimal solutions of N-Queens problems. The simulations show that the proposed algorithm is superior to other algorithms in light of successful rate, and it is especially suited to be used in practical system with parallel updating.