采用RBF和BP神经网络处理EMD分解过程中端点效应

在分析经验模态分解端点效应出现原因的基础上,采用BP和径向基函数神经网络预测法对端点效应进行研究.在实验中,通过延长信号的采样时间,使端点的数据延长,从而抑制EMD分解时产生的端点效应.同时为了比较两种数据延长方法的效果,分别将延长后的数据进行EMD分解.实验结果表明,这两种都可以有效抑制端点效应对分析结果产生的影响,提高经验模态分解的效果.

Using RBF and BP neural network processing EMD decomposition of end effects

On the analysis of empirical mode decomposition end effect appeared on the foundation of the reason,using BP and radial basis function neural network prediction method for the endpoint effect research.In the experiment,suppress the ending effect to inhibit end effect of EMD decomposed,by extending the signal sampling time.At the same time in order to compare two data extension method effect,respectively,will extend the data after EMD decomposition,which will extend the data after EMD decomposition.The experimental results show that,the two can effectively inhibit the endpoint effect on analysis results of impact,improved empirical mode decomposition effect.