基于改进EMD的爆破振动信号降噪方法  

De-noising method of blasting vibration signal based on improved EMD

作　　者：刘金山 LIU Jin-shan(Hebei Jidong Construction Engineering Co.,Ltd.,Tangshan 063210,Hebei,China)

机构地区：[1]河北冀东建设工程有限公司,河北唐山063210

出　　处：《工程爆破》2024年第4期156-164,共9页Engineering Blasting

基　　金：河北峪耳崖黄金矿业有限责任公司科研基金资助项目(2022-05)。

摘　　要：针对EMD存在端点效应和去噪效果不佳的问题,提出了改进EMD爆破振动信号降噪方法。首先,采用SSA优化SVM的方法对爆破振动信号延拓,并经过EMD方法分解获得不同IMF分量,然后,通过计算不同IMF分量和爆破振动信号之间的相关性系数,划分为含有噪声的IMF分量和包含有效信号的IMF分量并采用方差贡献率校核,最后,采用小波包对含有噪声的IMF分量进行降噪,重构获得降噪后的爆破振动信号。通过构造仿真信号和实测信号并采用评价指标,验证了SSA优化SVM的延拓方法抑制EMD端点效应的可行性。以实测爆破振动信号为例,采用均方根误差、信噪比以及小波包能量谱验证了改进EMD爆破振动信号降噪效果。研究结果表明,与镜像延拓方法相比,SSA优化SVM的延拓方法抑制EMD端点效应的效果更好;与CEEMDAN-小波阈值、CEEMDAN以及EMD相比,改进EMD方法信噪比最大,均方根误差最小,信噪比分别提高了1.77、4.62、1.94 dB,且降噪后信号高频带能量减小,中低频带能量占比均有不同程度提高,说明改进EMD方法不仅具有良好的降噪效果,也可有效保留原始信号能量信息。An improved EMD blasting vibration signal noise reduction method is proposed for the problems of endpoint effect and poor denoising effect of EMD algorithm.Firstly,the blasting vibration signal is extended by SSA optimized SVM,and different IMF components are obtained by EMD decomposition,and then the correlation coefficients between the different IMF components and blasting vibration signals are calculated,divided into IMF components containing noise and IMF components containing effective signals,and calibrated by the variance contribution rate.Finally,the wavelet packet is used to reduce the noise of the IMF component containing noise,and the noise-reduced blasting vibration signal is reconstructed.By constructing the simulation signals and the measured signals and using the evaluation indexes,the feasibility of the delayed approach of SSA optimized SVM to suppress the EMD endpoint effect is verified.The noise reduction effect of the improved EMD blasting vibration signal is verified by using the root-mean-square error,signal-to-noise ratio,and wavelet packet energy spectrum of the measured blasting vibration signal as an example.The results show that compared with mirror extension method,the SSA optimization SVM method has the best effect of suppressing the endpoint effect of EMD;compared with CEEMDAN-wavelet thresholding,CEEMDAN and EMD,the improved EMD method has the largest signal-to-noise ratio and the smallest root-mean-square error,and the signal-to-noise ratio is improved by 1.77,4.62 and 1.94 dB,and the energy of high-frequency band of the signal is reduced after the noise reduction of this method,and the energy proportion of middle and low frequency band is improved to different degrees,which indicates that the improved EMD method not only has good noise reduction effect,but also can effectively retain the original signal energy information.

关 键 词：爆破振动信号 EMD 小波包 降噪 

分 类 号：TD235[矿业工程—矿井建设]