基于改进粒子群优化算法的新型小波神经网研究

A Novel Wavelet Neural Network Based on Improved Particle Swarm Optimization Algorithm

摘要: 本文提出了改进的粒子群优化算法（Improved Particle Swarm Optimization，IPSO）的新型BP小波神经网络，并且对非线性辨识问题进行了仿真实验．实验结果表明，基于改进的粒子群优化算法的BP小波网络不仅具有小波分析良好的局部特性以及神经网络的学习、分类能力，而且具有粒子群优化算法全局快速寻优的特点．与简单的粒子群优化算法相比，该方法在收敛性和稳定性方面都有了较明显的提高，验证了它的合理性和有效性．

Abstract: A new BP wavelet neural network(WNN) based on improved particle swarm optimization(IPSO) algorithm is presented,and simulation experiments are made for nonlinear model identification.The simulation results show that the novel BP wavelet neural network based on IPSO not only has the good local property of wavelet analysis and the learning and classification capability of artificial neural network,but also has the fast global searching advantage of particle swarm optimization(PSO) algorithm.Compared with BP WNN based on simple PSO,convergence speed and stability of the proposed method are greatly improved,thus validating its rationality and effectiveness.