Providing a cost-effective solution to the two-fold problems of assigning locomotives to trains and train scheduling is of high importance for most railway companies. Due to high cost of deadheading locomotives and tardiness of trains, solving these two problems simultaneously has a significant impact on decreasing transport expenses in Iranian railways. The locomotive assignment problem is to assign locomotives to a set of freight rakes in order to provide the hauling of all freight rakes while maintaining minimum possible movements of deadheading locomotives and minimum coupling delay of freight rakes. In Iranian railways, passenger trains run across the network according to a fixed schedule. Scheduling freight trains is to determine the departure time, arrival time and the movement sequence of freight trains during the intervals between passenger trains in the way that they do not interfere with passenger trains’ schedule and at the same time minimum delay of freight trains occurs. In this paper the problems of locomotive assignment and train timetabling have been solved in two phases. In the first phase, locomotives are assigned to freight rakes using a genetic algorithm. In the second phase, the best locomotive assignment solution which is the output of the first phase is selected and a lower bound for arrival time of freight trains on their destination is calculated. Then freight trains are scheduled using another genetic algorithm. Using the applied method, the results of 30 test problems of simultaneous locomotive assignment and freight-train scheduling are presented.
)