Railway scheduling system using repair-based approach

In this paper, we propose an approach for railway scheduling based on iterative repair, a technique that starts with a complete but possibly flawed schedule and searches through the space of possible repairs. Based on the observation that the decision of the track assignments of trains would influence thepe$onnance of the resulting schedule, we divide the scheduling process into two levels. In theJirst level, a route preprocessing algorithm based on local search techniques is developed to find the appropriate track assignments for each train. In the second level, an iterative repair algorithm based on tabu search techniques is applied to repair the flawed schedule. The search is guided by an earliest-conflict-first heuristic that attempts to repair the earliest constraint violation while minimizing the value of objective function. We tested this approach with a randomly generated instance and found that it can work in an effective manner.