基于自适应VMD-MPE算法的矿山爆破地震波信号降噪方法研究

Noise reduction method of mine blasting seismic wave signal based on adaptive VMD-MPE algorithm

实测矿山爆破地震波信号含有大量高频噪声,一定程度上掩盖了真实信号特征,不利于爆破有害效应分析。为了有效降低实测信号的噪声成分,提出了基于自适应VMD-MPE算法的矿山爆破地震波信号降噪方法。将原信号进行变分模态分解(intrinsic mode function,VMD)获得本征模态函数(variational mode decomposition,IMF),利用能量差参数ξ自适应确定模态数K,对IMF分量进行多尺度排列熵(multi-scale permutation entropy,MPE)的随机性检测,准确区分出真实IMF和噪声IMF,去除原信号中的噪声IMF以达到降噪目的。对3组实测矿山爆破地震波信号进行降噪处理,结果表明该模型能够较好地去除高频噪声,保留了信号真实成分;其降噪效果均优于EEMD-MPE、CEEMDAN-MPE算法,验证了自适应VMD-MPE降噪方法的有效性。

A measured mine blasting seismic wave signal contains a lot of high-frequency noises,they cover up characteristics of real signal to a certain extent,and are unfavorable to analysis of blasting harmful effects.Here,to effectively reduce noise components of measured signals,a noise reduction method of mine blasting seismic wave signal based on the adaptive VMD-MPE algorithm was proposed.The variational mode decomposition(VMD)was performed for the original signal to obtain its intrinsic mode functions(IMFs),and the energy difference parameterξwas used to adaptively determine the mode number K.The randomness detection of multi-scale permutation entropy(MPE)was performed for IMFs to correctly distinguish real IMFs and noise IMFs,and remove noise IMFs in the original signal for the purpose of noise reduction.Noise reduction processing was done for 3 groups of measured mine blasting seismic wave signals,the results showed that the proposed method can better remove high-frequency noises and retain real components of the original signal;its noise reduction effect is better than those of EEMD-MPE and CEEMDAN-MPE algorithms to verify the effectiveness of the adaptive VMD-MPE algorithm.