基于改进EMD-小波包的爆破振动信号降噪方法研究  

Noise reduction method for blasting vibration signals based on improved EMD-wavelet packet

作　　者：闫鹏 张云鹏 侯善营 张为为 杨曦 YAN Peng;ZHANG Yunpeng;HOU Shanying;ZHANG Weiwei;YANG Xi(College of Mining Engineering,North China University of Science and Technology,Tangshan 063210,China;Hebei Provincial Key Lab of Mine Development and Safety Technology,North China University of Science and Technology,Tangshan 063210,China;College of Civil and Architectural Engineering,North China University of Science and Technology,Tangshan 063210,China)

机构地区：[1]华北理工大学矿业工程学院,河北唐山063210 [2]华北理工大学河北省矿业开发与安全技术重点实验室,河北唐山063210 [3]华北理工大学建筑工程学院,河北唐山063210

出　　处：《振动与冲击》2024年第11期264-271,287,共9页Journal of Vibration and Shock

基　　金：河北省教育厅在读研究生创新能力培养资助项目(CXZZBS2023124);河北省高等学校科学技术研究项目(编号:QN2023166);河北省自然科学基金(E2016209388);国家自然科学基金资助项目(52074124)。

摘　　要：针对经验模态分解(empirical mode decomposition, EMD)存在模态混叠和降噪效果不佳的问题,依据分解—正交—聚类—降噪—重构的思想,提出了改进EMD-小波包的爆破振动信号降噪方法。该方法融合了核主成分分析的正交性、K-means算法的聚类特性以及小波包的降噪优势,不仅可以消除EMD的模态混叠,也具有良好的降噪效果。研究结果表明:与自适应噪声完备集合经验模态分解(complete ensemble empirical mode decomposition with adaptive noise analysis, CEEMDAN)和EMD方法相比,在模拟信号降噪试验中,改进EMD-小波包方法的信噪比(7.9 dB)最大,均方根误差(2.96)最小。在实测爆破振动信号降噪中,改进EMD-小波包方法降噪后的信号与原始信号相关系数最大为0.91。改进EMD-小波包和CEEMDAN方法的降噪效果相对理想,且改进EMD-小波包方法对10~60 Hz低频信号能量保存效果较好,对60 Hz以上中高频噪声的滤除效果最好。Here,aiming at problems of mode mixing and poor denoising effect in empirical mode decomposition(EMD),according to the idea of decomposition-orthogonal-clustering-noise reduction-reconstruction,an improved EMD-wavelet packet noise reduction method for blasting vibration signals was proposed.This method could combine the orthogonality of kernel principal component analysis(KPCA),clustering characteristics of K-means algorithm and noise reduction advantages of wavelet packet.It could not only eliminate modal aliasing in EMD,but also have good noise reduction effect.The study results showed that compared with complete ensemble empirical mode decomposition with adaptive noise analysis(CEEMDAN)and EMD methods,in simulated signals noise reduction tests,improved EMD-wavelet packet method has the highest signal-to-noise ratio of 7.9 dB and the lowest RMS error of 2.96;in noise reduction of actually measured blasting vibration signals,the maximum correlation coefficient between signals after noise reduction with improved EMD-wavelet packet method and the original signals is 0.91;noise reduction effects of improved EMD-wavelet packet and CEEMDAN method are relatively ideal;improved EMD-wavelet packet method has a better energy preserving effect on low-frequency signals within the range of 10-60 Hz,and the best filtration effect on medium and high-frequency noise within the range of above 60 Hz.

关 键 词：爆破振动信号 经验模态分解(EMD) 核主成分分析(KPCA) K-MEANS算法 小波包 降噪 

分 类 号：TD235[矿业工程—矿井建设] O384[理学—流体力学]