Study on blasting data processing based on EEMD and wavelet threshold method
With the development of urban construction,the number of blasting events has gradually increased,and its data records are superimposed on the seismic waveform,which cause significant interference. Taking the ML 2.0 blasting event recorded on the Mudanjiang seismic station in Heilongjiang Province on June 14,2016 as an example,the vertical data record of the very wideband CTS-1EF seismometer was selected to perform the IMF decomposition and signal reconstruction of the blasting signal,and the signal-to-noise ratio (SNR) and correlation coefficient R were used as test indicators to evaluate the reconstruction effect. The results show that the ensemble empirical mode EEMD and wavelet threshold method are used for data preprocessing,the signal-to-noise ratio is larger,the correlation is closer to 1,and the reconstruction effect is better than the result of the single method. 8 ML 2.3-2.8 blasting events recorded by the Jingpo Lake Volcano Monitoring Network were selected,and the IMF decomposition and data reconstruction were carried out by the combined method of EEMD and wavelet decomposition,which results showed that the high-frequency noise interference was effectively reduced,the local main signal was highlighted,and the random noise was effectively suppressed. This method is used for noise reduction,and the reconstructed signal can better reflect the changing characteristics of blasting data,which can provide a reference for noise reduction analysis of blasting data.
万方数据
维普
