基于组稀疏约束的微地震震源参数谱投影梯度反演

震源参数反演是微地震监测中的关键技术,常规走时或逆时定位方法可以快速获取震源的空间位置,但是会忽略震源的时间信息.全波形反演(FWI)是一种有效的工具,利用完整的波形信息,通过选用合适的优化算法对微地震事件震源参数进行迭代反演,虽然存在计算量大的问题,但是反演出的结果信息丰富并且精度较高.本文依据微地震震源的特点,提出了基于谱投影梯度组稀疏约束的优化算法来进行震源参数全波形反演,模型测试结果表明: 该算法相比于逆时定位定位精度更高,且可以反演子波波形信息; 对低信噪比微地震记录具有一定的鲁棒性; 对不同时刻的多震源参数反演也能得到较好的结果; 该方法对速度模型具有敏感性,通过微地震数据更新速度模型再进行震源参数反演可以提高反演准确性.

Group sparse constrained inversion of microseismic sources parameters based on spectral projection gradient method

Source parameters inversion is the key technology in microseismic monitoring. Conventional methods including travel time and reverse time location can quickly obtain the source location,but they ignore the time information of seismic sources. Full waveform inversion method is an effective tool,which uses the complete waveform information to iteratively update the source parameters of microseismic events by selecting appropriate optimization algorithms. Although there is a large amount of calculation,the inversion results have rich information and high accuracy. According to the characteristics of the microseismic source,we develop an optimization algorithm based on the group sparse constraint of the spectral projection gradient for source parameters full waveform inversion. Numerical examples show that the algorithm has higher accuracy than reverse time location,and it can invert the source wavelet information. The full waveform inversion algorithm is robust to noise and can accurately calculate the parameters of multiple sources. The source parameters inversion method is sensitive to the velocity model. Therefore,we recommend using microseismic data to update the velocity model first,and then performing source parameters inversion to improve the inversion accuracy.