基于遗传退火进化算法的瞬变电磁非线性反演  

TRANSIENT ELECTROMAGNETIC NONLINEAR INVERSION BASED ON GENETIC ANNEALING EVOLUTIONARY ALGORITHM

作　　者：司元雷[1,2] 郑龙金 Si Yuanlei;Zheng Longjin(School of Resources and Geosciences,China University of Mining and Technology,Xuzhou 221116,Jiangsu,China;School of Information and Electronics Engineering,Jiangsu Vocational Institute of Architectural Technology,Xuzhou 221116,Jiangsu,China;The 5th Geological Team of Jiangxi Nonferrous Geology Prospecting Bureau,Jiujiang 332000,Jiangxi,China)

机构地区：[1]中国矿业大学资源与地球科学学院,江苏徐州221116 [2]江苏建筑职业技术学院信电工程学院,江苏徐州221116 [3]江西有色地质勘查五队,江西九江332000

出　　处：《计算机应用与软件》2020年第11期172-177,259,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(41674133)。

摘　　要：鉴于遗传算法的优秀全局搜索性能和模拟退火算法的局部搜索能力强的特点,提出遗传退火进化算法,其为模拟退火算法及遗传算法融合后形成的一种新的算法。它继承了两种算法各自的优点,避开了模拟退火算法易丢失全局信息以及遗传算法局部搜索能力差的缺点。使用Schaffer函数及Alpine函数对遗传退火进化算法进行数值实验,并对瞬变电磁理论地电模型进行反演试算。数值实验结果表明,该算法在寻优过程中更趋于避开早熟收敛,全局收敛性能得到改善,进而提高了反演精度,是一种可应用于地球物理非线性反演的有效方法。In view of the excellent global search performance of genetic algorithm and the strong local search ability of simulated annealing algorithm,genetic annealing evolutionary algorithm is proposed,which is a new algorithm formed by the fusion of simulated annealing algorithm and genetic algorithm.It inherits the advantages of the two algorithms,and avoids the shortcomings of the simulated annealing algorithm which is easy to lose global information and shortcomings of the genetic algorithm,which has poor local search ability.In this paper,Schaffer function and alpine function were used to carry out numerical experiments on genetic annealing evolutionary algorithm,and the inversion test of the transient electromagnetic theory geoelectric model was carried out.The results of numerical experiments show that the algorithm tends to avoid premature convergence and improve the global convergence performance in the process of optimization,thus improving the inversion accuracy.It is an effective method that can be applied to geophysical nonlinear inversion.

关 键 词：遗传算法 进化 模拟退火算法 非线性 反演 瞬变电磁 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] P631.325[自动化与计算机技术—计算机科学与技术]