基于矩阵变换的车辆路径问题仿真优化方法  

Simulation optimization method for solving vehicle routing problems based on matrix transformation

作　　者：胡祥培[1] 李永先[1,2] 郭建文[1] 

机构地区：[1]大连理工大学系统工程研究所,大连116024 [2]辽宁师范大学管理学院,大连116029

出　　处：《管理科学学报》2008年第4期67-75,共9页Journal of Management Sciences in China

基　　金：国家杰出青年科学基金资助项目(70725004);国家自然科学基金资助项目(70571009,70171040,70031020);教育部科学技术研究重点资助项目(03052);教育部博士点基金资助项目(20010141025);辽宁省教育厅基金资助项目(2004F020)

摘　　要：针对车辆路径问题这一求解难题,提出基于启发式变换的仿真优化原理和求解方法,建立了基于邻接矩阵的车辆路径问题的数学模型;利用启发式运行规则对仿真运行的参数进行了分析,通过矩阵变换改进优化搜索策略并找出最优解或满意解.算例求解表明,基于矩阵变换的仿真优化方法具有良好的稳定性和求解效率.该项研究为求解车辆路径问题这一难题提供了新思路.Focusing on the vehicle routing problems which are difficult to be solved, a principle and a method of simulation optimization with heuristic transformation are presented. A mathematical model of vehicle routing problems based on adjacency matrix is set up. The parameters produced by simulation are analyzed using heu- ristic rules. The matrix transformation is used to optimize searching strategies, and then the best solutions or ideal solutions are found. A case-study shows that the method of simulation optimization with matrix transfor- mation is efficient and its result has sound stability. This paper provides a new way for solving the vehicle routing problems.

关 键 词：车辆路径问题 仿真优化 邻接矩阵 矩阵变换 启发式变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]