多车场洒水车路径问题的双层遗传算法  

A Double-layer Genetic Algorithm for Sprinkle Car Routing Problem Based on Multiple Depots

作　　者：朱征宇[1] 刘建辉[1] 杨永[1] 谢志华[1] 

机构地区：[1]重庆大学计算机学院

出　　处：《微型电脑应用》2008年第6期1-3,共3页Microcomputer Applications

基　　金：高等学校博士学科点专项科研基金资助课题(20030611016);重庆大学骨干教师计划基金项目(2003A33)

摘　　要：论文提出了一种新的遗传算法对有多个加水点的洒水车服务路线问题进行优化求解,给出了一种多车场车辆弧路径问题的数学模型,并对传统遗传算法的染色体编码机制和种群结构进行了改进,设计了一种解决多车场车辆弧路径问题的双层遗传算法,可以表示出各车场出动的车辆及路径,与人工安排的方案进行比较,安排效率高,总行驶路程缩短15%以上,车辆行驶路线更为合理,有效地实现多车场车辆弧路径问题的优化。This paper presents a new genetic algorithm (GA) to solve the sprinkle car routing problem based on multiple depots and proposes a mathematical model of multiple-depot Vehicle Arc Routing Problem (MDVARP). In this paper, we improve the chromosome coding mode and population structure of the traditional genetic algorithm and devise a double-layer genetic algorithm to solve the MDVARP. Using the new GA, the vehicles and routes in different depots can be assigned. Compared with the manual assignments, the effectiveness of our algorithm is higher, the whole driving distances can be saved by 15 percent above, and the vehicle driving routes are more reasonable, the optimization of MDVARP can be realized effectively.

关 键 词：多车场 弧路径 车辆路径 遗传算法 洒水车 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]