Print Email Facebook Twitter Algorithms for Scheduling of Train Maintenance Title Algorithms for Scheduling of Train Maintenance Author Evers, R.P. Contributor Witteveen, C. (mentor) Huisman, B. (mentor) Faculty Electrical Engineering, Mathematics and Computer Science Department Software Technology Programme Algorithmics Group Date 2011-01-28 Abstract In this thesis we will develop algorithms for the scheduling of train maintenance at NedTrain facilities. We present a detailed analysis of the maintenance process at NedTrain. The problem of scheduling train maintenance is formalized and expressed using linear inequalities. We show that the Simple Temporal Problem can be used to find schedules that satisfy the temporal constraints of the scheduling problem. However, when resource constraints are added, the STP is no longer sufficient. The problem can still be expressed using the Disjunctive Temporal Problem but this cannot be efficiently solved. Two algorithms are presented: a zero/one integer linear programming approach and an STP-based approach. In the STP-based approach we start with an STP that describes the temporal constraints. The STP is then iteratively updated until its earliest start time assignment corresponds to a valid solution to the scheduling problem. Experiments show that the ILP method is fast enough to produce optimal base schedules whereas the heuristic method is well suited to updating schedules with little disruption when changes have occurred. Subject schedulingresourcesstpstnilp To reference this document use: http://resolver.tudelft.nl/uuid:f488ea95-f560-46f3-869e-66c6e095103c Part of collection Student theses Document type master thesis Rights (c) 2011 Evers, R.P. Files PDF thesis_ronald_evers_final.pdf 3.26 MB Close viewer /islandora/object/uuid:f488ea95-f560-46f3-869e-66c6e095103c/datastream/OBJ/view