Entropic-Based Robust Vehicle Rental Revenue Management with Substitution and Repositioning
针对车辆租赁行业订单随机到达、取消和租期不确定等挑战，提出一种基于熵风险度量的稳健方法，联合优化替代服务和库存重新定位，在多种风险偏好下最大化收益。基于真实数据的实验表明，该方法比期望优化基准模型平均周收入提高16%。
This paper investigates the practical inventory management challenges faced by the vehicle rental industry, which arise from the random arrival of orders and lead times before demand occurs, as well as the uncertainty and dynamic nature of cancellations and rental durations after demand is realized. To ensure on-demand rental services while mitigating the impact of randomness on inventory stability, the industry usually employs substitution service and inventory repositioning. We propose a tractable robust approach featuring an entropic risk measure. This approach jointly optimizes substitution service and inventory repositioning, addressing the maximization of revenue under varied risk-aversion preferences while safeguarding against multiple uncertainties and misspecification. Numerous numerical experiments constructed by real-world data exhibit that our model achieves 16% higher average weekly revenues compared with the expectation optimization benchmark model. Compared with the sample average approximation model, which disregards the variant nature of probability transitions, the model can achieve higher revenue with fewer vehicle inventories. Besides, in contrast to repositioning, our model is inclined to enhance the system’s revenue robustness by increasing the substitution ratio. History: Accepted by Pascal Van Hentenryck, Area Editor for Computational Modeling: Methods & Analysis. Funding: C. Fu is supported by the National Natural Science Foundation of China [Grants 72401229, 72310107003, 72271201]. N. Zhu is supported by the National Natural Science Foundation of China [Grant 72471163] and the Humanities and Social Science Fund of Ministry of Education of China [Grant 25JZD038]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.1086 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.1086 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .