A novel train rescheduling algorithm for correcting disrupted train operations in a dense urban environment

Abstract

Train rescheduling during disrupted service is a substantially significant task for urban railway operators. This task typically depends on the experiences and personal decisions of the professional operators. The operators use neither systematic methodologies for modifying train schedules nor quantitative criteria for measuring the quality of the rescheduled plans. Thus, operators have requested assistance in the form of a computer-aided train-rescheduling program. The authors have created a computer-aided train rescheduling system that seeks to minimize passenger inconvenience. The rescheduling algorithm calculates a value for the amount of inconvenience experienced by the passenger. In addition to a conventional passenger flow analysis, the authors algorithm considers passenger overflow, which is defined as waiting time experienced by a passenger while queuing on a platform. The proposed rescheduling method also introduces the idea that the time required to create, broadcast, and implement a new plan of train operation is itself a delay which must be accounted for in order to render a realistic passenger flow analysis during disordered operation. The causality of train system events after the initial event that caused the disruption of service is discussed for a realistic analysis of passenger flow.