Contact
Research Summary
An optimisation approach for railway network recovery actions in response to disruptions
The current practice to choose a recovery action after a disruption is based on pre-written contingency plans and, to some degree, relies on controller judgement without the ability to explore a variety of options automatically. Such complex decisions need to consider several different factors, such as the location of the disruption on the network, the current timetable, the level of perturbation in the network, etc. To ensure an effective and quick recovery plan, the controller needs to make a decision about which trains should be cancelled, rerouted, terminated short of the planned terminus, or regulated in a different way to minimise the disruption. The best decision is the one which delivers the least impact on train performance, and this can be measured in different ways, e.g. the amount of reactionary delay, cancellations or on time failures, or the count of trains or passengers affected. There is a clear need to be able to explore this large variety of options and factors in an automated way for a better-informed decision-making process to recover after a disruption.
The project will focus on developing a methodology to investigate the best recovery actions in response to a disruption. The methodology will explore a range of decision options based on the properties of the disruption and priority criteria of the controller. The ambition of the project is how to evaluate a range of best strategies and support the decision-making process.