讲座题目：基于稀疏-密集航班复制的航空公司巡航速度控制综合恢复交互机制

讲座专家：蒋建林

讲座时间：2025年9月14日（周日）上午9:00

讲座地点：天府校区雁归楼3017

报告摘要：Aircraft rerouting and passenger reallocation are critical in airline recovery. The recovery decision of aircraft rerouting greatly affects other recovery phases, and the recovery performance of passenger reallocation influences the direct revenue of airline companies and the social reputation of the airline industry. We consider aircraft rerouting and passenger reallocation from an integration perspective to preserve their interdependence, which is neglected by the regular sequential recovery. In addition, we take the cruise speed control into consideration to improve the recovery performance. As a common way to describe the adjustments of flights in disruptions, the flight copy approach requires a lot of flight copies to obtain high-quality recovery plans, especially when considering cruise speed control. Then, the network scale of the associated problem is quite huge, and thus, solving such an integrated problem is challenging due to the real-time requirement in disruption management. This paper introduces a sparse-dense flight copy approach and proposes an interactive mechanism that alternately adjusts aircraft routes on the sparse flight copy-based network and reallocates passenger itineraries on the dense flight copy-based network. These two networks become much smaller than those from the regular flight copy approach, and the solution difficulty decreases significantly. To realize such a mechanism, we develop a mathematical model to formulate the integrated recovery problem and propose a customized Benders decomposition (CBD) method to solve the model. The Benders decomposition strategy divides the proposed model into the aircraft rerouting subproblem and passenger reallocation subproblem, benefitting from the sparse-dense flight copy approach. By exploiting the properties and structure of the integrated recovery problem, we propose some acceleration techniques (including an effective feasibility certificate, scale management, and valid inequalities) to further promote the efficiency of the CBD method. Computational experiments on real-world data show that high-quality integrated recovery solutions (with optimality gaps within 5%) can be obtained by the proposed method within reasonable runtime. The effectiveness of acceleration techniques is also verified by the experiments.

专家简介：蒋建林，南京航空航天大学教授，博士生导师。研究方向为数值最优化、航空交通运筹与优化等。在国内外正式刊物上发表学术论文40余篇。主持国家自然科学基金重点项目1项、面上项目3项。获聘湖北省“楚天学者”特聘教授，入选江苏省“333高层次人才工程”培养对象、江苏省“青蓝工程”培养对象。