物流配送车辆路径问题多代竞争遗传算法.pdfVIP

2 5 Vol2 No5 2005 10 JOURNAL OF RAILWAY CIENCEAND ENGINEERING Oct. 2005 1 1 2 , , ( 1. 中南大学 交通运输工程学院, 湖南长沙410075; 2. 中国交通运输协会, 北京100053) : 针对遗传算法在解决车辆路 问题时效率较低的缺点, 将寿命和年龄的概念引入遗传算法, 提出了多代竞争遗传 算法每次通过遗传或变异更新 一定数量的 个体, 对于 个体按其年龄和寿命决定其是否留在下 一代种群中, 或终止其生 命通过这种方式, 增加了较优秀 个体在种群中的存活时间, 加大了其繁殖几率和优秀基因被子代 个体继承的概率并应 用算法对实际算例进行了测算, 取得较满意结果多代竞争遗传算法对其他优化问题同样适用 : 物流配送; 车辆路 ; 多代竞争; 遗传算法 : O221; U11 : A : 1672- 7029( 2005) 05- 0075- 05 Multi- generation compete genetic algorithms for logistics distribution vehicle routing problem 1 1 2 DENG Lianbo , SH Feng , MO Huihui ( 1. Traffic and Transportat ion Engineering School , Central South University, Changsha, 4 10075, China; 2. Chinese Communication and Transportation Association, Beijing 100053, China) Abstract: Age and life- span is introduced to genetic algorithms for overcoming the low efficiency, which is used to VRP. The multi- generation compete genetic algorithms is put forward. At each iteration, some individuals are updat ed by descendiblity and variation. Each individual, according to its age and life- span, is judged whether it is kept down to next generation population, or its life to be terminated. By this means, saved generations of some better indi viduals are prolonged in the generation population, and the probability of their propagat at ion and better- gene genetic increased. n the end, some better numerical solutions are presented through using the algorithms. Multi- Generation GA can also be used to solute the other optimal problems. Key words: logist ics distribution; vehicle rout ing problem; multi- generation compete; g