An improved BP NN for Nonlinear System Identification Based on PSO

摘要

Nonlinear system identification is to control engineering problems to be solved.Based on the deep analysis about BP neural network and particle swarm optimization,this paper puts forward the advantage in particle swarm optimization solving multi-object optimization problems will be introduced to the neural network training,provides a solution to optimization problem of the compromise of the accuracy and fastness of the neural network.Simulation experiments achieve greater precise identification results,and verify the effectiveness of the algorithm.