地球物理资料非线性反演方法讲座(八)——量子遗传算法  

Lecture on Non-Linear Inverse Methods in Geophysical Data(8) Quantum Genetic Algorithm and Its Applicaton in Geophysical Inversion

作　　者：罗红明[1] 王家映[2] 师学明[2] 朱培民[2] 

机构地区：[1]川庆钻探工程有限公司地球物理勘探公司,成都610213 [2]中国地质大学地球物理与空间信息学院,武汉430074

出　　处：《工程地球物理学报》2008年第6期635-642,共8页Chinese Journal of Engineering Geophysics

摘　　要：虽然线性反演理论目前已经相当成熟,但由于其方法本身比较依赖初始模型,而且容易陷入局部极小,在实际应用中常常显得"力不从心"。量子遗传算法QGA(Quantum Genetic Algorithm)以量子理论为基础,通过量子位编码和量子旋转门更新种群来寻找全局最优,加快了搜索速度,具有更强的全局寻优能力。通过对量子遗传算法内在机理的分析表明,QGA的寻优质量和效果明显优于传统遗传算法。This paper presents a new inversion algorithm, quantum genetic algorithm, which adopts the quhit chromosomes as presentations and updates the population using quantum rotation gate, to accelerate the search speed, to improve convergent efficiency, and to get a better global optimization. Having tested some synthetic and real data set, this paper strongly shows that the optimization quality and efficiency of QGA are better than the traditional genetic algorithm.

关 键 词：量子遗传算法 地球物理反演 全局寻优 遗传算法 

分 类 号：P631[天文地球—地质矿产勘探]