摘要
论文运用双种群遗传算法求解带软时间窗的旅行商问题,通过加入带有时间窗约束条件的惩罚函数,初始化两个种群,分别选择不同的交叉、变异概率。每次迭代后,交换种群间的优势个体所携带的遗传信息,以打破种群内的平衡状态,跳出局部最优解。双种群遗传算法比标准遗传算法显著提高了全局收敛性能。实验结果比较显示,该算法行之有效,具有较好的性能。
This paper puts forward a double populations genetic algorithm for traveling salesman problem with soft time windows,from joining a castigatory function with time windows restriction,initializing double populations,choicing the different intersect,aberrance probability each.After each iterative operation,the inherited information from preponderant unit of population is changed,so as to smash the population balanceable state,dap the local excellent explaination. Double populations genetic algorithm improves comprehensive convergent capability.This result shows that the procedure is efficient,which is proved by the number experiment provided by this paper.
出处
《计算机工程与应用》
CSCD
北大核心
2005年第33期74-75,78,共3页
Computer Engineering and Applications
关键词
旅行商问题
软时间窗
遗传算法
traveling salesman problem,soft time windows,Genetic Algorithm
作者简介
E-mail：qrwang2002@126．com
