关键词: 车辆路径问题；多车场；多约束；客户优先级；自适应的最大-最小蚁群算法

Abstract:

In order to save logistic distribution costs,a multi-depots vehicle routing problem with multiple constraints（MDVRPMC）was proposed.At first, a novel mathematical model to address the complicated issue of MDVRPMC was developed,the model took the clients’ priority levels,traffic condition influences,vehicle categories,time windows,and capacity constraints into consideration.And then,an adaptive max-min ant colony algorithm (A-MMACA) was proposed based on the combination of adaptive methods and ant colony algorithm with the maximal-minimal pheromone limit. A-MMACA possessed the ability to manipulate the pheromone updating process flexibly and expand the search scope.These characteristics keep A-MMACA away from the drawbacks of conventional ant colony algorithm (i.e., prematurity,local optimization,slow convergence speed).Finally,
a case study was presented to compare A-MMACA with Tabu-search algorithm and conventional ant colony optimum algorithm.The test results indicate that the proposed algorithm has more advantages than that of the fore mentioned algorithm in vehicle number,route length,route time,and computational speed for MDVRPMC.

Key words: vehicle routing problem;multi-depots;multiple constraint, clients&rsquo, priority level;adaptive max-min ant colony algorithm