基于多目标蚁狮优化算法的微震震源定位数学模型组合  

Mathematical Model Combination in Micro-Seismic Source Localization Based on Multi-Objective Ant Lion Optimization

作　　者：陈国庆 庞聪[3,4] 向涯 周正松 陈健 赵天文[1] CHEN Guoqing;PANG Cong;XIANG Ya;ZHOU Zhengsong;CHEN Jian;ZHAO Tianwen(Mathematical Modeling Research Center,Chengdu Jincheng College,1 Xiyuan Road,Chengdu 611731,China;College of General Education,Chengdu Jincheng College,1 Xiyuan Road,Chengdu 611731,China;Institute of Seismology,CEA,40 Hongshance Road,Wuhan 430071,China;National Observation and Research Station for Gravitation and Solid Earth Tides,40 Hongshance Road,Wuhan 430071,China;School of Computer Science and Artificial Intelligence,Wuhan University of Technology,122 Luoshi Road,Wuhan 430070,China)

机构地区：[1]成都锦城学院数学建模研究中心,成都市611731 [2]成都锦城学院通识教育学院,成都市611731 [3]中国地震局地震研究所,武汉市430071 [4]引力与固体潮国家野外科学观测研究站,武汉市430071 [5]武汉理工大学计算机与人工智能学院,武汉市430070

出　　处：《大地测量与地球动力学》2023年第10期1074-1079,共6页Journal of Geodesy and Geodynamics

基　　金：四川省软科学研究项目(23RKX0351);中国地震局地震研究所和应急管理部国家自然灾害防治研究院基本科研业务费(IS202236328);武汉引力与固体潮国家野外科学观测研究站开放基金(WHYWZ202208)。

摘　　要：通过多目标智能优化算法研究微震震源定位存在的模型组合合理性未阐明、易陷入局部最优解、定位结果波动性较大等问题。为解决这些问题,首先在到时差模型与到时差商模型基础上设计4个不同的微震震源定位数学模型,两两组合构建6个多目标优化定位模型;再设计3组基于不同台网形状(三维多面体、二维长方形、一维直线型)的微震震源正演仿真实验和1组工程数据验证实验,并引入多目标蚁狮优化(multi-objective ant lion optimization,MOALO)算法求解这些模型;最后采用多个统计指标评判各个模型组合定位效果的优劣。结果表明,数学模型组合(TDA-P1,TDQA)结合MOALO算法的多目标优化定位策略能够得到较高的微震震源定位精度,且模型稳健性较好,优于其他模型组合和传统多目标定位方法,在微震监测领域具有一定的应用价值。There are problems in micro-seismic source localization using the multi-objective intelligent optimization algorithm:the rationality of model combination is not elucidated,it is easy to fall into local optimal solutions,and the large fluctuation of localization results.To solve these problems,we design four different mathematical models for micro-seismic source localization based on the arrival time difference model and the arrival time difference quotient model,and six multi-objective optimization localization models are constructed by combining two and two.We design three sets of micro-seismic source forward simulation experiments and one set of engineering verification experiments based on different network shapes(3D polyhedral,2D rectangular,and 1D linear),and introduce multi-objective ant lion optimization(MOALO)algorithm.We apply several statistical indicators to evaluate the advantages and disadvantages of the localization effect of each model combination.The results show that the combination of mathematical models(TDA-P1,TDQA)and the MOALO algorithm can obtain high accuracy of micro-seismic source localization,and the robustness of the models is better than other model combinations and traditional multi-objective localization methods,which has certain application value in the field of micro-seismic monitoring.

关 键 词：微震震源定位 多目标优化 数学模型组合 到时差商模型 到时差模型 蚁狮优化算法 

分 类 号：P313[天文地球—固体地球物理学] P315[天文地球—地球物理学]