Nonlinear System Identification Based on Radial Basis Function Neural Network Using Improved Particle Swarm Optimization

Abstract

A novel method of nonlinear system identification based on constructing radial basis function neural network using particle swarm optimization algorithm with mutation operator is proposed. After determination of units of number in RBF layer, all parameters in relevant network such as central position, spreading constant, weights and offsets of RBF NN are coded to particles in learning algorithm. The parameter vector, which has a best adaptation value, is searched globally. By the comparison with standard particle swarm optimization algorithm, the simulation results show the effectiveness of this method.