基于到时时空分布的微震震源定位算法

Micro-seismic source location algorithm based on spatio-temporal distribution of arrival times

为了提高微震震源定位精度和算法鲁棒性,研究了基于到时的定位算法的目标函数构建和时空区域信息使用。仿真实验基于Numpy,采用理想到时+随机扰动来模拟生产情况。结果表明:到时定位算法中常用的L_(1)和L_(2)目标函数对实际震源存在偏离且偏离现象非常普遍。为了减轻伪震源问题,使用到时偏差的似然函数作为基础函数,聚合时空邻域的单点基础函数作为目标函数;新构建的4维区域分布算法(Distribution of 4-Dimention Area Algorithm,D4DA)在2次标定炮定位中将定位误差分别降低了14%和26%。

In order to improve the precision of the micro-seismic hypocenter location and the robustness of algorithms,the objective function construction and spatiotemporal region information use of the location algorithm based on time are studied.The simulation experiment is based on Numpy,the perfect arrival times subjoining random disturbance is used to simulate production situation.Result shows that the L_(1) and L_(2) objective functions deviate from the hypocenter and the deviation is quite common.To alleviate the pseudo source problem,the likelihood function of arrival times deviation is used as the basic function,and the information of space-time neighborhood is then combined as the new objective function.The distribution of 4-dimension area algorithm(D4DA)reduces the positioning errors by 14%and 26%respectively in the two calibration gun positioning.