Prognostics-based Scheduling to Extend a Distributed Platform Production Horizon under Service Constraint: Model, Complexity and Resolution.
Résumé
In the field of production scheduling, this paper addresses the problem of optimizing the useful life of a heterogeneous distributed platform composed of identical parallel machines and which has to provide a given production service. Each machine is supposed to be able to provide several throughputs corresponding to different operating conditions. The purpose is to provide a production scheduling that maximizes the production horizon. The use of Prognostics and Health Management (PHM) results in the form of Remaining Useful Life (RUL) allows to adapt the schedule to the wear and tear of machines. This work comes within the scope of Prognostics Decision Making (DM). The key point is to configure the platform, i.e., to select the appropriate profile for each machine during the whole production horizon so as to reach a total throughput based on a customer demand as long as possible. In the homogeneous case, the Longest Remaining Useful Life first algorithm (LRUL) is proposed to find a solution and its optimality is proven. The NP-Completeness of the general case is then shown. A Binary Integer Linear Programming (BILP) model which allows to find optimal solutions for fixed time horizons has been defined. As solving such a BILP is NP-Complete, solutions can however be computed in reasonable time only for small size instances of the problem. Many heuristics are then proposed to cope with large scale decision problems and are compared through simulation results. Exhaustive simulations assess the efficiency of these heuristics. Distance to the theoretical maximal value comes indeed close to 5% for the most efficient ones.
Domaines
Automatique / Robotique
Fichier principal
ac3149eb-0475-48fe-a512-1fac73df2f4c-author.pdf (923.06 Ko)
Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...