基于BP神经网络的函数逼近的MATLAB实现

Realization of Function Approximation on Matlab Based on BP Neural Network

BP神经网络由于它的学习能力和非线性特性,使其能够对非线性函数进行很好的逼近。通过对BP神经网络结构和MATLAB软件及其BP神经网络工具箱的应用研究,利用BP神经网络工具箱设计BP神经网络,用于对非线性函数的逼近,通过网络的训练、测试达到了预期的效果。

Owing to its learning ability and nonlinear characteristics, BP neural network has the strong ability to approximate functions. Based on the research on the structure of BP neural network and the application of Maflab software and BP neural network toolbox, this paper uses BP neural network toolbox to design BP neural network to approximate nonlinear functions. Desired effect is achieved through the training and testing of network.