Multi-mode Resource Constrained Project Scheduling Problem with resource flexibility
Résumé
We discuss the problem of project scheduling encountered in an assembly enterprise. This problem requires an integration of resource limitation, worker skills, and multiple possible execution modes for each task. The duration of each activity depends on the amount of resources allocated to its execution, and the objective is to minimize the overall project duration. This problem is modeled as a multi-mode resource constrained project scheduling problem (MRCPSP) in the context of skilled worker and resource flexibility. Because the basic MRCPSP is combinatorial and belongs to the class of NP-hard problems, exact procedures are limited in the ability to solve big size problem in a reasonable computational time. In this paper three approaches are presented. The first method is a parallel scheduling heuristic, called MMSH, which enumerates at each time decision all the non-dominated schedulable activity-mode combinations and schedule the one with the best performance. The second method is a tabu search (TS) algorithm with a powerful neighborhood function based on a flow graph representation in order to implement various search strategies. The search of the solution space is carried out via two types of move strategy. Furthermore, the TS algorithm is embedded in a decomposition based heuristic (DBH) which serves to reduce the solution space. The last proposed method is an hybridization of Tabu Search and MMSH, this later gives an excellent results. The effectiveness of the methods is demonstrated through extensive experimentation on different standard and new benchmarks instances and proves that our results are competitive.