【摘要】 针对现场采集的水轮机水导摆度信号中出现的异常尖峰成分,利用EEMD进行分解处理,分离出不同成分的特征信号,除去尖峰异常特征成分以及杂波成分的本征模函数(Intrinsic Mode Function,IMF)分量。将剩余的IMF分量进行信号重构,从而获得了除去异常尖峰成分后的重组信号。通过对仿真的信号进行分析,验证了该方法对水轮机摆度信号特征提取的有效性。更多还原

【Abstract】 A characteristics extraction method based on the ensemble empirical mode decomposition( EEMD) is proposed.The basic principle and solving procedures of the method are introduced. The abnormal peak constituents of the collected hydraulic turbine swing signal was decomposed by EEMD,and the characteristic signal with different constituents was separated. The abnormal peak characteristic constituents and the clutter constituents of IMF were eliminated,and then the rest of IMF was reconstructed to obtain the restructured signal. Through the analysis of the simulation signal,it has been proven that the method for extracting the characteristics of hydraulic turbine swing signal is effective.更多还原