基于本征模态函数最优配比的冲击波信号经验模态分解降噪方法

摘要

The original signal was decomposed with EMD method,got the best ratio of the different IMFs though system calibration and ergodic search and combined them to get the de-noised signal. In sweep frequency signal simulation experiment.The result eliminated the noise and kept the part of high frequency.In blast wave experiment,the result eliminated the noise and also kept the part of high frequency in the front edge.The relative error of the peak pressure,the time of positive pressure and the impulse are only-4.91%,1.06% and 2.12%,better than the result of wavelet de-noise method.%将原始信号进行EMD分解,通过系统校准和遍历寻优求得各个IMF的最优配比因子,将其重构后得到降噪信号.扫频信号仿真实验,结果去除了系统噪声,且对信号高频分量保留完好.爆炸冲击波实测实验,降噪结果有效去除噪声,且较好地保留冲击波前沿高频分量.在峰值压力、正压作用时间和冲量3个方面与原始信号偏差仅为-4.91%、1.06%和2.12%,优于小波降噪结果.