报告题目：Generalized Riskiness Index in Vehicle Routing under Uncertain Travel Times: Formulations, Properties, and Exact Solution Framework

时间：2024年11月25日上午9:00-10:00

地点：中关村校区主楼216

报告人：张真真

报告人简介：

张真真，同济大学经济与伟德国际1946bv官网副教授、博士生导师。入选上海市高层次人才计划。长期致力于大规模整数规划和不确定优化的理论研究与算法设计，及在物流与运输规划、智能制造等方面的应用。目前已发表高质量论文30余篇，包括Operations Research、INFORMS Journal on Computing、Transportation Science、Transportation Research Part B、NeurIPS等，主持国家自然科学基金青年项目及优秀青年项目、上海市人才项目和华为、中远海运科研课题各1项，创新研究群体项目“综合运输系统运营管理”骨干成员。现任管理科学与工程学会交通运输分会执行秘书长、世界交通大会货运与物流系统优化技术委员会委员、运筹学会随机服务与运作管理分会理事，并长期担任Operations Research，Transportation Science等30多个国际知名期刊的审稿人。

报告内容简介：

We consider a vehicle routing problem with time windows under uncertain travel times where the goal is to determine routes for a fleet of homogeneous vehicles to arrive at the locations of customers within their stipulated time windows to the maximum extent while ensuring that the total travel cost does not exceed a prescribed budget. Specifically, a novel performance measure that accounts for the riskiness associated with late arrivals at the customers, called the generalized riskiness index (GRI), is optimized. The GRI covers several existing riskiness indices as special cases and generates new ones. We demonstrate its salient managerial and computational properties to motivate it better. We propose alternative set partitioning-based models of the problem. To obtain the optimal solution, we develop an exact solution framework combining route enumeration and branch-price-and-cut algorithms, in which the GRI is dealt with in route enumeration and column generation subproblems. We mainly reduce the solution space by exploiting the GRI and budget constraints’ properties without losing optimality. The proposed method is tested on a collection of instances derived from the literature. The results show that a new instance of the GRI outperforms several existing riskiness indices in mitigating lateness. The exact method can solve instances with up to 100 nodes to optimality. It can consistently solve instances involving up to 50 nodes, outperforming state-of-the-art methods by more than doubling the manageable instance size.

（承办：管理工程系、科研与学术交流中心、中国运筹学会数据科学与运筹智能分会（筹））