BP网络的Matlab实现及应用研究

Realization and Applications of BP Network on Matlab

人工神经网络以其具有信息的分布存储、并行处理以及自学习能力等优点,已经在信息处理、模式识别、智能控制及系统建模等领域得到越来越广泛的应用。他的基于误差反向传播算法的多层前馈网络,即BP网络在非线性建模、函数逼近和模式识别中有广泛的应用,介绍了BP网络的基本原理,分析了Matlab人工神经网络工具箱中有关BP网络的工具函数,并给出了部分重要工具函数的实际应用。

Artificial neural network is widely used in the fields of information processing ,pattern recognition ,intelligent control and system modeling,which has advantages of information distribute saving,parallel processing and self -teaching ability. Its multiple - layer feedforward network based on back propagation arithmetic, namely BP networks, are widely applied in nonlinear modeling, function approach and pattern recognition. This paper introduces the fundmental of BP network ,analyzes tool functions of BP network in the toolbox of ANN in Matlab,and demonstrates applications of some important tool functions.