DI-UMONS : Dépôt institutionnel de l’université de Mons

Recherche transversale
(titres de publication, de périodique et noms de colloque inclus)
2015-06-06 - Travail avec promoteur/TFE - Anglais - 60 page(s)

Equeter Lucas , "Optimization of an opportunistic maintenance policy subject to resource constraints", Dehombreux Pierre (p) , Fleurquin Guillaume, Letot Christophe, 2015-06-06

  • Codes CREF : Sciences de l'ingénieur (DI2000), Probabilités (DI1132), Mécanique (DI1240), Mécanique appliquée générale (DI2100), Statistique appliquée (DI1133)
  • Jury : Kouroussis Georges (p)
  • Unités de recherche UMONS : Génie Mécanique (F707)
Texte intégral :

Abstract(s) :

(Anglais) In an era of cost-optimized industrial decisions, the field of maintenance is no exception and is subject to a series of researches, be it for optimal preventive maintenance grouping or simply new maintenance policies. Among the new practices, opportunistic maintenance sees every system stoppage as an opportunity for preventive maintenance. The objective of this final year’s thesis is to produce a comprehensive maintenance policy that includes corrective, preventive and opportunistic maintenance actions for any series system. It proposes a new opportunistic criterion and studies the influence of some factors, in particular a limited maintenance crew situation. The analyze focuses on the global daily cost of maintenance and the implications on the workload of the repairmen crew. This final year’s thesis falls in line with previous researches in the field of opportunistic maintenance optimization. The series system that is studied in this work is moreover subject to failure modes that are grouped together in so-called failure types, that model the need of repairmen of different skills to perform an action on specific parts of machines. In this work, the preventive maintenance is performed on an age-based system, and the opportunistic maintenance possibility is evaluated both on corrective and preventive actions, in hope for a natural preventive maintenance grouping occurring through the opportunistic process. This thesis is carried out using a Scilab program that uses the Monte Carlo method to produce an important failure times sample and simulate the system behavior over time. This program is intended to be able to function with any number of machines in a series system, themselves subject to any number of failure modes. It also includes resource management (workforce and tangible resources) and restrictions. A plethora of criteria triggering opportunistic actions is available, and this thesis focused on the reliability of failure types as such criterion. A variety of simulations were run with varying parameters, and in particular cost ratios between preventive and corrective maintenance actions. The model developed in this thesis demonstrates that a proper opportunistic policy may allow to optimize the global cost of maintenance. It also shows the influence that the maintenance policy has on the repairmen workload. Obviously, every industrial system being different, no generic conclusion can be drawn from the results, but the model proves worthy in opportunistic trigger optimization and repairmen crew dimensioning under all situations for series systems, as it allows to distinguish optima regarding these factors.