提出了一种解决位置管理问题的差分进化算法,给出了一种将采用浮点编码的种群个体映射为问题解的方法、基于问题特性的种群初始化启发式方法,以及早熟收敛问题的解决策略。基于随机生成的数据对算法进行了模拟实验,将该算法的结果与遗传算法、禁忌搜索算法及蚁群算法进行了对比。
A differential evolution algorithm for loca-tion management problem is proposed. A mapping method from population individual of float coding to problem solution is put forward. An effective heuristic population initialization technology based on problem characteristic and solutions to overcome the premature convergence problem of the algorithm are studied. The proposed algorithm is simulated with random produced input data. The experiment result of the proposed algorithm is compared with the results of genetic algorithm, Tabu search algorithm and ant colony algorithm.