No abstract

No abstract

In manufacturing systems, mixed model assembly lines are used to produce different products to deal with the problem of customers’ demands variety, and minimizing the cycle time in such assembly line is a critical problem. This paper addresses the mixed model assembly line balancing problem type 2 that consists in finding the optimal cycle time for a given number of workstations.  A hybrid Greedy randomized adaptive search procedure-Genetic algorithm is proposed to find the optimal assignment of tasks among workstations that minimize the cycle. A Ranked Positional Weight heuristic is used in the construction phase of the proposed GRASP, and in the local search phase, a neighborhood search procedure is used to ameliorate the constructed solutions in the construction phase. The GRASP is executed many times in order to seed the initial population of the proposed genetic algorithm, and the results of the executions are compared with the final solutions obtained by the hybrid GRASP-GA. In order to test the proposed approaches, a numerical example is used.

No abstract

Des industriels utilisent des lignes d'assemblage spéciales à modèles mixtes. Lors de la conception de ces lignes, un problème critique apparaît, celui de l'équilibrage, qui doit être résolu pour minimiser le nombre de postes de travail ou le temps de cycle. Une amélioration de la procédure de recherche adaptative gloutonne basée sur la variation de la liste candidats a été proposée.