基于遗传-拟牛顿混合算法的远场震源定位

Seismic source localization of the far field based on GA-mixed quasi-newton algorithm

在微震震源定位过程中,震源与检波器的相对位置对定位精度有较大影响。由于计算机计算的精度有限以及预置的速度误差,当震源位于远场时,定位精度会比较低。为了解决这个问题,在设定均匀介质的条件下,基于无需测速的震源定位模型,结合遗传算法全局优化及拟牛顿法的快速局部搜索能力,提出用遗传-拟牛顿混合算法震源进行定位。为了验证该混合算法的优势,用遗传-拟牛顿混合算法和简单遗传算法进行MATLAB仿真模拟对比,实验结果表明,使用该混合算法获得的震源定位速度更快、精度更高。

In the process of seismic location,the relative position between the source and detector has a great influence on positioning accuracy. Because of the limited computer calculation precision and preset speed error,when the source is placed on the far field,the location accuracy will be low. In order to solve this problem,in this paper,with the condition of setting homogeneous medium,based on the source localization model without speed,combined with the global optimization of genetic algorithm and fast local search ability of quasinewton method,the GA-Quasi-Newton mixed algorithm is proposed to find source localization. In order to validate the advantage of this hybrid algorithm,the GA-Quasi-Newton mixed algorithm is compared with simple genetic algorithm by MATLAB simulation. The experimental results show that the hybrid algorithm for source localization has faster speed and higher precision.