Mixed Model Assembly Line Sequencing to minimize delays using meta-heuristics

Abduh Sayid Albana (1) , Karim Aroui (1) , Gülgün Alpan (1) , Yannick Frein (1)

1 G-SCOP_GCSP - Gestion et Conduite des Systèmes de Production

Nowadays, the customers are increasingly demanding, pushing the companies to offer highly diversified products. This requires that different kinds of products be manufactured in intermixed product sequences on the same line. Such assembly lines are called the mixed-model assembly lines (MMAL). Workers and machinery have to be flexible to reduce the setup times and costs. A good vehicle sequence in MMAL can have many positive effects on MMAL: It can permit producing more products in a shorter time period (providing cost reductions), it can also improve the work conditions by balancing the workload of the operators. In this article, we are interested in the sequencing of a mixed model assembly line for Truck Industry. In the literature, different objectives exist to solve the MMAL sequencing problem. In this article, we present methods to minimize the total work overload. In a previous work, a linear programming (LP) approach has been proposed for this problem. The MMAL is known to be an NP-hard problem. The exact methods such as LP can only handle small problems and their applications are limited in an industrial context. Therefore, we present here solutions based meta-heuristics. Three types of meta-heuristic algorithms are used in this research: Genetic Algorithm (GA), Simulated Annealing (SA), and finally a hybrid method based on both Genetic Algorithm and Simulated Annealing (GASA). Numerical tests are carried out to compare the performance of the proposed algorithms. For small instances, a benchmark data from the literature is used to compare the performance of the meta-heuristics versus the optimal solution found by the LP approach, based on the computational time and the quality of the solutions. The comparisons are also made for larger instances, for some generated data and the data from an industrial case study.

Mots clés: Mixed Model Assembly Line; Sequencing Problem; Meta-heuristics

Domaines: Recherche opérationnelle [math.OC]

Abduh Sayid Albana, Karim Aroui, Gülgün Alpan, Yannick Frein. Mixed Model Assembly Line Sequencing to minimize delays using meta-heuristics. Joint International Symposium IMSS '14 and CIE '44, Oct 2014, Istanbul, Turkey. ⟨hal-01236967⟩