基于AWSST和GCV的矿山微震信号联合降噪新方法

A new denoising method for microseismic signal in mines based on AWSST and GCV

由于露天矿山开采环境的复杂性和各异性,传感器采集得到的微震信号包含大量的噪声干扰,为了获得更加准确的微震信号特征,提出了基于自适应同步压缩小波变换(Adaptive Synchrosqueezing Wavelet Transform,AWSST)和广义交叉验证(Generalized Cross-Validation,GCV)的联合降噪方法,从而对微震事件定位和判断开挖过程中岩体状态提供可靠的数据保障。首先在自适应小波变换(Adaptive Wavelet Transform,AWT)的基础上,对每个尺度中的小波系数进行阈值化处理,利用GCV方法自动确定每个成分的最佳阈值水平,达到去除噪声的目的。然后,通过同步压缩变换(Synchrosqueezing Transform,SST)的后处理操作,在时频平面对小波变换系数进行二次重分配,从而提升时频表达的能量聚集性。最后,应用于微震信号的降噪处理,并与现有的时频分析方法进行比较,结果表明提出的联合去噪方法降噪效果更好、时频分析的分辨率更高。

Due to the complexity and heterogeneity of the open-pit mining environment,the microseismic signal collected by the sensor contains a lot of noise.To obtain more accurate microseismic signal characteristics,this paper proposes a method based on Adaptive Synchrosqueezing Wavelet Transform(AWSST)and Generalized Cross-validation(GCV)denoising method,to provide a reliable data guarantee for locating microseismic events and judging the state of rock mass during excavation.Firstly,the researched method performs the thresholding of the wavelet coefficients in each scale based on Adaptive Wavelet Transform(AWT)and uses the GCV method to automatically determine the optimal thresholding level of each component.Therefore,the purpose of removing noisy components is achieved.Later,through the post-processing operation of Synchrosqueezing Transform(SST),the wavelet transform coefficients are redistributed twice in the time-frequency plane,thereby improving the energy concentration of time-frequency representations.Finally,this method is applied to denoising of microseismic signals and compared with the existing time-frequency analysis methods.The result demonstrated that the joint denoising method proposed in this paper has better denoising performance and higher resolution of the time-frequency plane.