Model and Solution Method for Railroad Crew and Vehicle Rescheduling

摘要

The crew rescheduling problem (CRP) and the vehicle rescheduling problem (VRP) are problems of reactively making schedules for a set of crews or vehicles in response to transport disorder and changes in train operations. We propose a 0-1 integer programming model for CRP/VRP based on the multicommodity flow network, and approximation method with heuristics and local search. The proposed model is used to formalize a "difference from the original schedule", which is a significant criterion in solving CRP/VRP. Experimental results to real-world data of vehicle rescheduling are also described.