Author(s)

Teruomi Katori & Takashi Izumi

Abstract

Railway transport has high punctuality, however, sometimes accidents happens due to facility failure or natural disasters, and the operation needs to change from its original scheduled timetable. When the operation has to return to the original schedule that operation is called “train rescheduling”. Automatic rescheduling methods have been suggested considering several operation conditions and operation environments.

During the rescheduling, the operated rolling stock types must be considered because different types of rolling stock have limitations for each running section, for example, one type of vehicle cannot run sections that are driven in a different electrical manner.

Based on the above, this paper describes how to generate automatic rescheduling methods while specifically considering the operated rolling stock types. We suggest a rescheduling method which does not use software tools such as mixed integer linear programming. Train rescheduling deals with the problem of how to move rolling stocks from their location after resumption time to their location at time of recovery to scheduled conditions. Thus considering the conditions, track routes are registered for each type, and the created rolling stock operation plan is a combination of the possible routes during rescheduling time. Other stock operation routes are also searched for, and the combination of all these routes results in a complete rescheduling plan. When more than one train is located at the same location at the same time, the solution is rejected because the plan is impossible to be realized.

Average headway time and its standard deviation at each station are used as evaluation functions to decide which solution to adopt from the multiple rescheduling timetables that can be generated. Shorter average headway time results in a higher train number, and accordingly more passengers can be transported. With a smaller deviation, trains run at similar intervals and the number of passengers is equalized for all operated trains. The operation plan with the smallest evaluation value is adopted as the final rescheduling timetable.

Finally, we apply the suggested method to a theoretical track modelled on an existing track for a variation of traffic accident parameters, and indicate that rescheduling timetables are generated that fulfil the operation conditions of the rolling stock types under consideration.

Keywords

average headway time, rescheduling plan, rolling stock types, route combinations, train operation.