改进的自适应遗传算法及其工程应用

Improved Adaptive Genetic Algorithm and Its Application to Backward Analysis of Geotechnical Engineering

引进小生境技术、种群迁移以及增加杂交个体之间的海明距离对自适应遗传算法进行了改进,从而建立了改进的自适应遗传算法,改善了传统的遗传算法局部收敛和早熟的现象,大大加快了全局搜索的速度以及搜索全局最优解的概率.工程实例表明:提出的改进自适应遗传算法应用于岩土工程的位移反分析具有搜索速度快、精度高等优点;同时对初始种群的形成方式、种群规模以及最大杂交概率、最大变异概率进行了参数分析.

By niching, population migration and increasing Hamming distance between different populations,an improved adaptive genetic algorithm is proposed. It keeps populations diversity effectively and improves the performance of local convergence and premature convergence and enhances search speed and precision of genetic algorithm. High search speed and precision of the proposed method were further tested through back analysis of geotechnical engineering, and the influence of initial population yielding method, quantity of populations, maximum crossover probability and maximum mutation were discussed on search speed and precision.