Abstract: Based on the forecasting and solving algorithm of the uncertain demand in the dynamic vehicle routing problem with uncertain demand information in the existing literature, a multidimensional data layer customer demand forecasting method and a proactive real-time control method are established to discuss the potential customer response criteria. The mathematical model of dynamic vehicle routing problem is established by introducing proactive real-time control method with the aim of minimizing the total transportation cost. The genetic algorithm is used to solve the model. The test results show that the proposed model and algorithm can respond to the customer’s demand more timely and effectively by applying the Joybuy’s distribution data of 20 clients in Chongqing area and the two-stage comprehensive proactive adjustment strategy.

Key words: dynamic vehicle routing problem, mutli-data layer structure, pro-active real-time control method, demand forecast

摘要： 通过对现有文献中需求信息不确定的动态车辆路径问题，在不确定需求预测和求解算法的基础上，建立了多维数据层客户需求预测方法和前摄性实时控制方法，讨论了潜在客户的响应准则。以总运输成本最小为目标，构建了引入前摄性实时控制方法求解动态车辆路径问题的数学模型，改进了遗传算法对该模型进行求解。应用京东在重庆地区的客户点的配送数据及两阶段综合前摄性调整策略，验证了设计算法的性能，实验结果表明设计的模型及算法可以对客户需求进行更及时有效的响应。

关键词: 动态车辆路径问题, 多维数据层结构, 前摄性实时控制, 需求预测