A Method of Train Regulation Planning by the Constraint Programming Using Train Regulation Knowledge

Abstract

Recently, train regulation is being complex because of congested train schedules and various kinds of cars/classes of trains. So it is necessary for decreasing operator's workload and raising his skill to help his work by computer systems. A research in automatic train rescheduling has been done by several approaches. But it is impossible to satisfy standards of train regulation about response time and accuracy. In this paper, we propose a concept of an automatic train regulation method using operator's knowledge and show an algorithm that is built by the constraint programming. Being evaluated on some test case, it is shown that the algorithm can find a feasible solution at practical speed.