针对目前人工信号识别技术和傅里叶变换在分析识别矿山微震信号时的局限性,提出Matlab小波工具箱分析方法。首先对矿山实际监测到的干扰矿震信号进行小波变换,对小波分解之后的矿震信号进行频谱分析,结果表明能够准确地观察到信号的突变点,确定矿震信号P波初到时,从而更加精确得出矿震震源的位置和能量大小;其次对受干扰矿震信号进行小波消噪,同时对小波分析的4种阈值降噪方法进行分析对比,发现无偏似然估计阈值效果最好,显示了小波分析的强大消噪功能。研究结果表明,小波分析具有良好的矿震信号识别效果和消噪能力,是矿震信号处理和分析的一种有效方法。

Due to the limitations of the present artificial signal recognition technology and Fourier transform in analyzing the mine microseismic signal, the Matlab wavelet toolbox analysis method was presented. Through the transformation of the microseismic signals monitored in mine, prior to the spectral analysis of transformation signals, the discontinuity of the signals were accurately observed and the first arrival time of P wave was determined. As a result, the accurate location and energy of mining shocks were concluded. Further, through the wavelet de-noising of signals and comparing the four kinds of de-noising methods, the unbiased estimate threshold method worked best and this showed the powerful function of the wavelet analysis in de-noising. Therefore, it is revealed that the wavelet analysis is an effective method of seismic signal processing and analysis as it is capable of seismic signal recognition and noise elimination.