摘要
研究了用小波神经网络对非线性系统进行建模问题。提出了用带遗忘因子的最小二乘法训练网络的权值 ,利用递推预报误差算法训练伸缩因子和平移因子的交互辩识算法。仿真结果证明了算法的有效性。
Wavelet neural networks are used to solve the problem of modelling of nonlinear system.Weights of the network are trained by the Least Square method with forgetting factors. And the scaling factor and shift factor are trained by using the error prediction algorithm .The simulation result has shown the effectiveness of proposed algorithm.
出处
《自动化技术与应用》
2004年第2期22-24,共3页
Techniques of Automation and Applications
