Optimizing train parking and shunting at NS service yards

More Info
expand_more

Abstract

This thesis proposes several methods to solve the planning of parking and servicing of trains in a shunting yard. Management of temporary superfluous resources and maintenance is a recurring topic in many fields. Trains visit a service yard regularly to be cleaned, maintained, and to store them outside rush hours. These activities are currently scheduled by hand, which is difficult and time consuming. This thesis proposes several heuristics to construct an assignment method to solve this problem, which is referred to as the extended Train Unit Shunting Problem. A discrete process simulation is constructed for two real life service yards, namely ‘Kleine Binckhorst’ and ‘Cartesiusweg’, to assess the methods. The methods are not capable of obtaining a perfect solution, but most proposed methods are able to solve problems of a reasonable size in a few seconds. The methods incorporate logistical knowledge obtained in practice to optimize flow through the yard, resulting in a shunting policy that is effective and easy to implement in daily practice.