Imperialist competitive algorithm for unrelated parallel machine scheduling with sequence-and-machine-dependent setups and compatibility and workload constraints
Milad Elyasi,Yagmur Selenay Selcuk,O. Örsan Özener,Elvin Coban
DOI: https://doi.org/10.1016/j.cie.2024.110086
IF: 7.18
2024-03-22
Computers & Industrial Engineering
Abstract:In this paper, we present an in-depth analysis of the unrelated parallel machine scheduling problem (UPMSP) in the context of washing machine production at Vestel Electronics. Vestel Electronics is a leading producer of washing machines and holds a noteworthy market position in the European consumer electronics industry. The production process at Vestel Electronics is make-to-order (MTO) and requires 20 assembly lines to produce 200 different products. The decision maker must consider several important factors while forming the production schedule, including job-assembly line compatibility, the release times and due dates of the jobs, and a workload balance among the different assembly lines. This study aims to develop an algorithm to minimize the total earliness and tardiness while considering sequence-and-machine-dependent setups, unequal release times, machine-job compatibility restrictions, and workload balance requirements. To address this complex scheduling problem, we propose a novel algorithm, the imperialist competitive algorithm (ICA), which has a strong exploration-exploitation balance while mimicking real-world dynamics. Our numerical results show that the ICA significantly outperforms the current practice in Vestel by 39% and the best-performing algorithms in the literature by 23% and 12%.
computer science, interdisciplinary applications,engineering, industrial