带有稳定学习算法的小波神经网络及应用

针对当系统存在未建模动态时,神经网络辨识易产生参数漂移和不稳定的问题,采用输入-状态稳定性(ISS,input-to-state stability)分析方法,获得小波神经网络权值矩阵和小波尺度参数的误差反传类时变学习算法,该算法不带有鲁棒修正即可以实现小波神经网络的鲁棒稳定性.仿真例子表明,此稳定学习算法优于一般的误差反传算法,并将带有稳定学习算法的小波神经网络用于污水处理过程出水水质COD(化学需氧量,chemical oxygen demand)的预测,获得了较好的效果.

Wavelet Neural Networks with Stable Learning Algorithm and Its Application

In the presence of unmodeled dynamics,the parameters' drift even instability may occur in the identification system of neural networks.Input-to-state stability(ISS)approach is applied to achieving the error backpropagation-like time-varying learning algorithm of weight matrix and wavelet scaling parameters in wavelet neural networks,of which the robust stability is guaranteed without robust modification.Simulations showed that the stable learning algorithm outperforms conventional error backpropagation ones...