Locomotive assignment and freight train scheduling using genetic algorithms

This study considers the movement of freight trains through a passenger rail network, a common occurrence in many developing countries. Passenger trains run according to a fixed schedule while freight trains need to be accommodated and run on the same track, ensuring that they do not interfere with passenger train movements. Operationally, this requires the assignment of a locomotive to a freight rake and then creating a workable schedule. Accordingly, we propose to solve the problem in two phases. In the first phase, we assign locomotives with partial scheduling with the objectives of minimizing total deadheading time and total coupling delay. We use a genetic algorithm to find non-dominant locomotive assignment solutions and propose a method for evaluating its performance. The solutions are then ranked using two approaches, based on the decision maker's preferences. In the second phase, we select a locomotive assignment solution based on the ranking and find the lower bound on the arrival time of freight trains at their destinations. We use a genetic algorithm again to schedule the freight trains in the passenger rail network, with prescribed locomotive assignment precedence constraints with the objective of minimizing total tardiness. Computational results confirm the efficacy of the proposed method.