A study on sorting strategies in marshaling yards with a limited number of tracks and track capacity

Abstract

In marshaling yards, freight wagons are sorted from inbound trains to outbound trains for further transport. To organize an efficient shunting process, sorting strategies are proposed in the literature. The application of sorting strategies is generally restricted by the number of classification tracks and their lengths. This can lead to difficult-to-implement or even inoperative sorting plans. To handle limited track capacity, we decompose the shunting process into a series of consecutive periods of time resembling timetables of inbound trains. A heuristic is used in every period to decide on the postponement of inbound trains when track capacity is scarce. This way, sorting strategies become applicable on a rolling time basis. A strategy is said to solve a shunting task when it enables building all outbound trains within a given time horizon. We examine the performance of five well-known sorting strategies for a large set of shunting tasks within a computational study. The simulation results indicate that the sorting strategies perform differently when numbers and lengths of classification tracks vary. In conclusion, we are able to determine the most reliable strategy among the set of considered sorting strategies for a marshaling yard of a certain size.