IMPACT OF INCENTIVE REGIMES AND UNCERTAINTY ON RAILWAY RESCHEDULING

Many models have been applied to railway rescheduling during disruption in recent years to assist in understanding the solution space. This research has included the development of methods that seek to minimise both operational change and passenger disruption.

These models have focussed on disruption resolution for time windows of 30 to 120 minutes due to computational restraints. However, given the nature of railway services, this duration does not allow for a full cycle of services, and therefore does consider the profile of passenger flows throughout the day. This limitation prevents representation of when rescheduled services are most needed to be available to ensure there is sufficient capacity for expected passenger flows. I am therefore interested in developing models with a longer time horizon which are able to consider when train capacity is most required during disruption.

In this project, I will develop a railway disruption model with the ability to handle longer time windows. The new model will minimise total passenger discomfort throughout the remaining day, whilst incorporating the uncertainty associated with the ending of the event which has caused disruption. This combination of passenger discomfort and uncertainty has not previously been undertaken.