基于改进经验模态分解法的地铁列车轮轨应变信号降噪

摘要

Wheel and rail strain signal acquisition is influ-enced by various factors,in which exist many problems like the random white noise and lower accuracy.For this reason,an improved EMD method is proposed to reduce the noise,en-hance the accuracy of local mean value solution and suppress the mode mixing phenomenon. The improved EMD method is used to analyze the measured signal of the wheel/rail force of metro trains. Research results show that the improved EMD noise reduction method can effectively eliminate the random interference and influence of white noise in wheel and rail strain signal,having great practical significance for the effec-tive identification of real wheel/rail force signal.%地铁列车轮轨应变信号的采集受到诸多因素影响,存在随机白噪声,准确性较差.对此,提出一种改进的经验模态分解法用以降噪,可提高局部均值求解精度,抑制模态混叠现象.运用该法对地铁列车轮轨应变实测信号进行分析,结果表明:改进经验模态分解法能有效消除轮轨应变信号中随机白噪声的干扰和影响,对有效识别轮轨力真实信号具有重要意义.