At the end of the day's operation the arriving train units should be parked at a depository while waiting to begin the tasks the following day. This problem is called the Shunting Problem. Planning which train units will cover the requested tasks is a complicated process and there are many factors to be included in the decision. Today this planning is conducted manually, which is a time-consuming process that requires a lot of experience and knowledge. In this thesis we have developed a prototype of a software program that executes the shunting process efficiently, while we minimize the maintenance costs of the regular inspections, which are required for each train unit. Minimizing the costs is based on an increased focus on the restrictions associated with the maintenance inspection. The prototype is designed as a decision support system to help out the operator to ensure a robust planning of the allocation of train units and the subsequent parking within a short computation time. We will focus on the coalition between the software and the user to achieve the best possible result. The efficiency of the prototype is tested by small test examples, where we demonstrate how the prototype presents different scenarios, and how the operator may affect the solution afterward if he finds it possible to optimize. At the end of this thesis we will demonstrate why our prototype is profitable and how it can ensure a reduction in overall maintenance costs by reducing the number of inspections on an annual basis. The interaction between the decision support system and the operator must ensure that the maintenance of the train units are conducted in a timely manner, to reach, but not exceed the amount of kilometers driven in correlation to the demands that DSB S-tog are assigned to.
|Educations||MSc in Mathematics , (Graduate Programme) Final Thesis|
|Number of pages||135|