论文部分内容阅读
本文基于小波网格系提出一种用非均匀分布样本训练MWN的新方法,避免了当样本分布不均匀时难以发挥MWN优点的缺陷;同时给出了该算法的逼近精度分析该方法的最大特点是:计算简单,便于在线应用最后用于辨识非线性动态系统,仿真结果验证了该方法的可行性和有效性。
In this paper, a new method of training MWN with non-uniform distribution samples is proposed based on wavelet grids. This method avoids the disadvantage of not being able to exert the advantages of MWN when the samples are unevenly distributed. At the same time, the maximum accuracy of the proposed method is analyzed. Is: Computation is simple and easy to use online Finally, it is used to identify nonlinear dynamic system. The simulation results verify the feasibility and effectiveness of the method.