遗传算法的改进及其在应力场反演中的应用  

Improvement of Genetic Algorithm and Its Application on Stress Field Inversion

作　　者：朱守彪[1] 

机构地区：[1]北京大学地空学院地球物理学系,北京100871

出　　处：《西北地震学报》2005年第2期97-100,共4页Northwestern Seismological Journal

基　　金：地震科学联合基金项目(103016);国家自然科学重点基金项目(40234042)

摘　　要：利用人工方法产生多样化的初始群体,引入“移民”机制并采用小种群搜索,运用自动调整交换概率与变异概率的方法将遗传算法进行了改进,改进后的遗传算法在防止早熟,提高收敛速度方面有一定改善。最后利用改进的遗传算法反演了青藏高原的应力场,其结果与地质结果有一定的相似性。Genetic Algorithm is improved in this paper,using production of initial population manually, introduction of immigration of population, searching in small population, adoption of adaptive probability of cross and mutation. In the example, the modified Genetic Algorithm has some improvement in preventing premature convergence and enhancing the searching efficiency. And finally, the stress field in Qinghai-Tibet Plateau is inversed with improved Genetic Algorithm, the result shows similar to geological one.

关 键 词：遗传算法 青藏高原 应力场 

分 类 号：P553[天文地球—构造地质学] O224[天文地球—地质学]