混合编码遗传算法及其在非线性反演中的应用  

HYBRID ENCODING GENETIC ALGORITHM AND ITS APPLICATIONIN SOLVING NON-LINEAR INVERSION

作　　者：陈超[1] 陈家联[1] 余丰[1] 

机构地区：[1]中国地质大学地球物理系,武汉430074

出　　处：《计算机应用与软件》2004年第3期16-18,69,共4页Computer Applications and Software

基　　金：科技部"九五"重点攻关项目第五课题 (96 - 91 4 - 0 5);湖北省自然科学基金项目 (0 2 0 650 1 2 );国土资源部矿产资源定量预测及勘查评价开放式重点实验室研究基金项目(MGMR2 0 0 1 - 9)资助

摘　　要：遗传算法中编码机制对交换和变异的搜索能力有重要影响。本文分析了二进制与十进制编码的搜索特点 ,提出了混合编码遗传算法的技术构想 ,并结合大变异技术 ,有效地提高搜索及产生有效基因物质的能力。论文结合地球物理反演问题的非线性特点 ,应用混合编码遗传算法求解地球物理位场反演问题 。Encoding system of genetic algorithm determines the searching abilities of genic operations.This paper present a Hybrid Encoding Genetic Algorithm (HEGA),which is based on two encoding mechanism.In the procedure the higher mutating probability should be taken in order to create more chances for the production of new genes.After analyzing the genetic operations and their encoding procedures,it is easy to find out it is effective method to mix the two encoding into the algorithm and it may be helpful to solve the inversion of geophysical data.The paper introduces the theoretical computation of synthetic models and the application of gravity inversion of 2D model by using HEGA.

关 键 词：混合编码遗传算法 非线性反演 参数编码 理论模型 二进制编码 

分 类 号：O242.23[理学—计算数学]