地面阵列式微地震数据盲源分离去噪方法

The blind separation denoising method for surface array micro-seismic data

由于地面噪声干扰多,原始微地震监测数据的信噪比相对较低,而数据品质决定了微地震事件的定位精度,因此,提高监测数据信噪比是微地震处理的关键环节。充分利用地面阵列式微地震数据监测台站数量多、间距近和分布广的优点,采用基于互相关的盲源分离去噪方法,进行随机噪声的压制。基于互相关盲源分离的阵列式微地震数据去噪方法,采用负熵作为目标函数,以粒子群优化方法进行高效求解,并通过互相关系数有效解决了盲源分离的不确定性问题,实现了地面阵列式微地震事件分量与噪声干扰分量的有效分离。通过正演模拟数据和实际微地震监测数据的去噪效果分析,证实了该方法的有效性和实用性。

Due to the ground noise interference,the signal-to-noise ratio(S/N) of original micro seismic monitoring data is relatively low,and the data quality determines the positioning accuracy of micro seismic events.Therefore,it is very important to improve the S/N of monitoring data in the processing of micro-seismic event.Based on the advantages of more monitoring sites,short intervals and wide distribution in ground array,the method of blind source separation based on cross-correlation method can denoise random noise.Blind source separation method is based on cross-correlation method,which uses negentropy as the objective function,with particle swarm optimization method for high efficiency solving.Through the cross correlation coefficient to solve the uncertainty problem of blind source separation,the method can realize effective separation of the ground array micro seismic events components and noise interference components.Through the forward simulation signal and the actual ground array micro-seismic data processing,the method was effectively applied to enhancing S/N of the ground array micro seismic data,which confirms that the method is feasible and effective.