Optimization of disposition in case of disruptions from a passenger's view

该项目致力于研究公共交通延误和扰乱对乘客的影响，并提出如何在运营处置中作出反应的建议。虽然我们已经为面向乘客的等待决策处理奠定了基础，但我们项目的下一阶段将解决方法和算法的开发，这些方法和算法可以支持更复杂的处理用例(例如，取消、部分取消、短期周转、绕道、轨道变更)。为了在实际中应用，这些算法的实时运行是至关重要的，由于公共交通需求的不断增加和高峰时段的高载客率，可用车辆容量扮演着重要的角色。在我们的项目中，我们将开发新的方法，在配置中考虑车辆容量。目标是引导乘客，从而避免过度拥挤的车辆，并尽可能平衡车辆占用。开发能够处理不可靠信息的方法仍然是在中断情况下处理的主要挑战。在我们以前的工作中，我们已经证明了部署决策的优化建议强烈地依赖于假设的客流。因此，我们的目标是借助额外的数据来源更准确地估计当前的客流。异质数据源的融合及其动态更新带来了数据集成的问题，需要解决。将扩大和改进稳健性措施，以便量化中断和处置对客流的影响。特别是，我们将研究针对空间和时间的动态需求而优化的时间表的稳健性。

This project is devoted to the impact of delays and disturbances on passengers in public transport and it develops recommendations how to react in operational disposition. While we have built the basis for a passenger-oriented disposition of waiting decisions, the next phase of our project addresses the development of methods and algorithms which can support more complex use cases of disposition (for example, cancellations, partial cancellations, short turnarounds, detours, track changes). To be applicable in practice, it is crucial that these algorithms run in real-time.Due to continuously increasing demand in public transport and high occupancy during rush hours, available vehicle capacities play an important role. In our project, we will develop new methods which consider vehicle capacities within disposition. The goal is to guide passengers so that overcrowded vehicles can be avoided and vehicle occupancy is as balanced as possible. The development of methods which are capable to cope with unreliable information remains a central challenge for disposition in case of disruptions. In our previous work we have shown that optimized recommendations for disposition decisions strongly depend on the supposed passenger flow. Therefore, we aim at more precise estimations of current passenger flows with the help of additional data sources. The fusion of heterogeneous data sources and their dynamic update leads to problems of data integration which have to be solved. Robustness measures which allow for a quantification of the impact of disruptions and disposition on passenger flows will be expanded and refined. In particular, we will study the robustness of timetables which have been optimized with respect to dynamic demand in space and time.