基于车公里成本的多车型车辆规划方法

Multi-model Vehicle Scheduling Based on Per Vehicle-kilometer Cost

多车型车辆路径规划是车辆路径问题(VRP)的难点,在求解VRP前先进行多车型规划能降低其求解难度,针对基于车辆车公里成本的多车型规划又涉及VRP求解的问题,把车辆路径分为仓库点到分区中心路径和分区中心到客户之间的路径,构建了只与客户间总路径相关的基于车公里成本的多车型规划模型,避免求解每辆车的行驶路径;根据随机分布旅行商问题(TSP)路径估算模型主要与客户的地理分布离散度有关,利用相关分析建立了客户间平均路径与离散度的关系,把客户之间的路径转化为客户间的平均路径;利用分枝定界法求解多车型规划模型,并以实际配送数据进行验证实验。结果表明:基于成本的多车型规划比基于装载率的配送成本平均降低10%-25%。

For the multi-model vehicle scheduling problem based on per vehicle-kilometer cost which also involves VRP, we divide the vehicle route into the section from the warehouse point to the regional center and the section from the regional center to the customer, and construct the multi-model vehicle scheduling model based on per vehicle-kilometer cost only related to the combined sections from the customers instead of the driving distance of each vehicle. In view that the stochastic distributed traveling salesman problem(TSP) route estimation model is mainly related to the geographical dispersion of the customers, we have a correlation analysis to establish the relation between the average route and the dispersion of the customers, and convert the route between customers into the average route between them.Next, we use the branch and bound method to solve the multi-model vehicle scheduling model. At the end, through a verification experiment with actual distribution data, we demonstrate that the cost-based multi-model vehicle scheduling could reduce the average distribution cost by 10%-25% compared to the load rate based scheduling.