关键词: 移动充电车, 动态预约需求, 实时响应策略, 近似动态规划

Abstract: Real-time response problem for demand of mobile charging vehicles under appointment mode is new in the mobile charging industry. This problem contains two types of demands with different characteristics and dynamic alternating influence, which not only havetime window constraints and real-time response requirements but also havedynamic uncertainty. In order to solve this problem, with the goal of maximizing the overall income, we propose a linkage two-stage strategy, which uses approximate dynamic programming to solve the future value and is integrated into the two stages of demand response: In the first stage, a dynamically updatable service scheme is generated by combining the multi-stage stochastic dynamic decision model and the Tabu search algorithm. In the second stage, a real-time response decision-making process for dynamic arrival demand is proposed based on the first stage. Finally, the comparative experiment verifies the effectiveness of this strategy under different customer sizes and degrees of dynamism, as well as obtains the managerial implications. This study can provide decision support for the formulation of real-time demand response strategy for mobile charging vehicles, and has enlightening significance for similar demand response strategy problems with dynamic characteristics.

Key words: mobile charging vehicles, dynamic appointment demands, real-time response strategy, approximate dynamic programming