基于粒子群算法的岩体微震源分层定位方法

 针对微震经典定位法速度模型给不准和联合定位法震源位置、发震时间和微震传播速度相互关联，解不唯一的问题，提出一种微震源定位分层处理方法，先对微震信号进行消噪、分波、到时修正和劣质信号剔除等一系列处理，初步获取正确的有效信号；然后以相邻两传感器监测到时之差与计算到时之差的残差平方和最小为目标，利用粒子群算法，识别微震源位置和速度模型；接着，根据识别到的微震源位置和速度模型，以传感器监测到时和计算到时的残差平方和最小为目标，直接求解微震源发震时间的解析解；最后，再次结合矿山实际开采现状反分析有效微震信号选取的正确性和微震源定位的准确性，必要时再次对微震信号进行处理和定位，较好地解决经典法速度模型给不准和联合法解不唯一的问题。与经典法相比提高了微震定位精度，与联合法相比提高收敛速度和解的稳定性；关联性分析表明某些震源坐标在使用分层法定位时和速度具有一定的关联性，并给出震源坐标和速度相互关联的必要条件和相互关联的几个特殊位置；算法性能分析表明为了进一步提高算法的收敛速度、定位精度和解的稳定性，传感器布置要尽量：(1) 使重点关注的区域位于传感器阵列之内，且距离传感器尽量近，(2) 避免可能发生微震的震源处于能使震源坐标和速度相互关联的位置上。现场爆破试验进一步验证微震源分层定位方法的可行性；最后讨论几种速度模型的选取，分析几种速度模型的优劣及工程应用的可能性。

MICROSEISM SOURCE LOCATION WITH HIERARCHICAL STRATEGY BASED ON PARTICLE SWARM OPTIMIZATION

According to the fact that it is difficult for typical microseism location(TMSL) algorithm to select a proper velocity structure and that it is not easy for joint microseism location(JMSL) algorithm that velocity structure，microseismic source coordinates and time of microseismic occurrence are associated with each other to identify unique solution，the intelligence microseism source location algorithm with hierarchical strategy(IMSHL) is suggested. Firstly，effective microseism signals are acquired by removing background noise，dividing P-wave，S-wave，correcting monitored arrival time，deleting poor microseism signals and so on；secondly，velocity structure and microseismic source coordinates are identified using particle swarm optimization(PSO) based on minimizing residual sum of squares of calculated and monitored difference of arrival time between two neighboring sensors；thirdly，analytic solution of microseism occurrence time can be gained by minimizing residual sum of squares of calculated and monitored arrival time based on the identified velocity structure and microseismic source；finally，the correction of selecting microseism signals and the accuracy of microseism source location are verified by back analysis based on the in-situ state of practical exploitation，and if necessary，source location and identification of microseism signals should be carried out for the second time. A simple example shows that convergence precision of the algorithm is improved compared with TMSL，and convergence velocity and the stability of the algorithm are also improved compared with JMSL. The correlation of coordinates of certain seismic sources and velocity for IMSHL is shown by correlation analysis，and its necessary condition and locations of several special seismic source are also provided. The analysis of the performance of IMSHL shows that，in order to further improve convergence velocity，precision and stability of solution，microseismic source should be in ranges of sensor array and close to the sensors as much as possible by proper disposal of the sensors；and the sensors should not be in the location that makes coordinates of seismic sources and velocity associated with each other. Lastly，the feasibility and advantages of IMSHL are verified by an in-situ blasting test and the applicability of several velocity structures is discussed and analyzed.