摘要： 提出了一种基于改进遗传算法（Improved Genetic Algorithm , IGA）的新型BP小波神经网络，并通过异或问题和非线性辨识问题进行仿真实验。实验结果表明，基于改进遗传算法的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，而且具有遗传算法全局快速寻优的特点，与简单遗传算法相比，在收敛快速性和稳定性方面都有了明显的改善。

关键词: 神经网络, 小波分析, 遗传算法, 系统辨识

Abstract: A new BP wavelet neural network based on improved genetic algorithm (IGA) is proposed, and the simulation experiments about the XOR problem and the nonlinear-model identification have been done. The simulation results show that the new BP wavelet neural network based on improved genetic algorithm not only has the local property of wavelet analysis and the learning and classification capability of artificial neural network, but also has the advantages of the fast global searching of genetic algorithm. And compared with BP WNN based on simple genetic algorithm (SGA), it is clearly improved in terms of convergent speed and stability.

Key words: Neural network, Wavelet analysis, Genetic algorithm, System identification

中图分类号: 

• TP18