WIT Press

Maintenance Plan Optimization For A Train Fleet


Free (open access)





Page Range

349 - 358




204 kb

Paper DOI



WIT Press


K. Doganay & M. Bohlin


Maintenance planning is an important problem for railways, as well as other application domains that employ machinerywith expensive replacements and high downtime costs. In a previous paper, we have developed methods for efficiently finding optimized maintenance schedules for a single unit, and proposed that the maintenance plan should be continuously re-optimized based on the condition of components. However, fleet-level resources, such as the availability of expensive spare parts, have largely been ignored. In this paper, we extend our previous approach by proposing a solution for the fleet level maintenance scheduling problem with spare parts optimization. The new solution is based on a mixed integer linear programming formulation of the problem. We demonstrate the merits of our approach by optimizing instances of maintenance schedules based on maintenance data from railway companies operating in Sweden. Keywords: maintenance planning, condition based maintenance, optimization, mixed integer programming, railways. 1 Introduction Maintenance planning is an important issue, especially for application areas where high cost machinery is used, and when time spent on maintenance disrupts the operation and causes losses, monetary or otherwise. Industry often fears that introducing condition based maintenance (CBM) will lead to more frequent service interventions, which could counter the potential value of implementing CBM. Implementation should therefore be done with care, as the maintenance planning process under CBM needs to be adapted to a much more dynamic situation. We have previously [1] proposed to harvest the full potential value in CBM for rail vehicle maintenance using a combination of condition monitoring and onlinemaintenance planning.A side effect of using this dynamic approach, instead


maintenance planning, condition based maintenance, optimization, mixed integer programming, railways