Passenger-oriented Disruption Management in Railways

Subject

Disruptions are inevitable in the daily operation of the Netherlands railways due to some unforeseen events such as rolling stock breakdown, switch failure, signal failure, etc. Currently, the way to deal with disruptions is to use the pre-defined contingency plan that is usually not optimal or even of low quality as the disruption scenarios are quite various. Besides, such a contingency plan is usually operation-oriented without taking passengers into account. Therefore, this research will be developing methods to rapidly respond to different disruption scenarios with rescheduling solutions that are optimized from the perspective of passengers.

Scientific challenges

One of the scientific challenges is to develop an agent-based dynamic passenger assignment model under uncertain disruption length, capable of reflecting each passenger’s response well. Another challenge is to integrate the passenger assignment model with the rescheduling part to form a passenger-oriented rescheduling model. Since the disruption is under a real-time environment, it is necessary to develop an algorithm that can find a passenger-oriented rescheduling solution with acceptable quality within the required time period.

Societal relevance

This research focuses on the development of models and algorithms to solve the passenger-oriented rescheduling problem in case of disruptions. On one hand, it can be a good assistance for the dispatchers to propose effective dispatching strategies. On the other hand, it can improve the service level of the railway system to guarantee passengers’ conveniences even in the disrupted circumstances.

Yongqiu Zhu

Start/end date: 11th April, 2016- April, 2020
Daily supervisor: Rob M.P. Goverde
Promotor: Serge Hoogendoorn