为改善遗传算法局部寻优精度较差的固有缺陷,该文引入一种新的基于遗传算法的模式搜索法求解地下水管理模型.首先通过在遗传算法中引入精英保留的进化策略,提高遗传算法的收敛性,进行全局搜索得到全局近似最优解,在此基础上采用局部寻优类算法模式搜索法进一步改善优化结果,最后将该混合优化方法用于求解一个地下水管理的基准问题.计算结果表明,基于改进遗传算法的模式搜索法,改善了单一遗传算法局部寻优精度较差的缺陷,在地下水管理模型的优化求解中具有较好的应用前景.%In order to overcome the inherent disadvantage of the genetic algorithm, an improved hybrid optimization method was introduced to solve groundwater management models. The improved hybrid method was based on genetic algorithm and a local search algorithm named pattern search method. Firstly, the elitist strategy was introduced into GA to improve the global convergence, and an approximate global optimal solution was obtained;secondly, the pattern search method was utilized for further improving the approximate global optimal solution; finally, the introduced hybrid optimization method was adopted to solve a benchmark groundwater management problem. The simulation-optimization results reveal that this hybrid op timization method, to some degree,can overcome the disadvantage of the genetic algorithm,and may have a good application perspective on solving groundwater management model.
展开▼
