Algorithms for multi-stage optimization problems: Integrated optimization for public transport

本课题以公交综合规划为例，对多阶段优化问题进行了分析。我们考虑站点位置、线路规划、时刻表和车辆调度。我们将综合规划过程作为一个整体进行分析。重点将放在综合优化中的稳健规划上。一个目标是分析在顺序过程的哪个阶段最好考虑健壮性问题。第二个重点是研究如何使模型和过程适用于真实世界的实例。我们处理了现实客运路线模型的集成，并考虑了车辆的通行能力和乘客的第一英里和最后一英里。我们计划开发一种基于我们在本项目第一部分中开发的特征模型的实用方法。使用分解方法，我们进一步研究了允许同时计划高峰和非高峰时段的算法。展望未来的发展，例如移动应用程序，我们还想讨论是否仍然需要周期性时间表的问题。我们计划调查哪些线路应该定期提供服务，哪些线路可以使用呼叫或类似程序的小型车辆提供服务。所有结果都将在基准实例上进行测试和评估。

This project analyzes multi-stage optimization problems with emphasis on integrated planning in public transportation as test case. We consider stop location, line planning, timetabling, and vehicle scheduling. We analyze the integrated planning process as a whole. The focus will be on robust planning in integrated optimization. One goal is to analyze in which stage of a sequential process robustness issues are best to be considered.A second focus delivers research on making models and procedures applicable for real-world instances. We deal with the integration of realistic passenger routing models, and we take capacities of vehicles and the first and last mile of the passengers into account. We plan to develop a pragmatic approach that is based on the Eigenmodel which we developed in the first part of this project. Using decomposition approaches we furthermore work on algorithms which allow simultaneous planning for peak and non-peak hours.Looking into future developments, such as mobile apps, we also want to discuss the question if periodicity of schedules is still needed. We plan to investigate which lines should be served with a periodic schedule and which can be served by small vehicles using dial-a-ride or similar procedures.All results will be tested and evaluated on benchmark instances.