COMPUTERS & INDUSTRIAL ENGINEERING, vol.173, 2022 (SCI-Expanded)
In this study, lot streaming in hybrid flowshop scheduling problem under machine capability and limited waiting time (lwt) constraints is investigated. The HFSP plays a vital role in the effective management of a multi-stage manufacturing system. To the best knowledge of the authors, the problem with its distinctive characteristics is addressed for the first time in the literature. A novel generic optimization model with the makespan objective is formulated to represent the problem mathematically. Structural properties and a lower bound formulation are presented to further explore the problem. Since the problem belongs to the NP-hard class, the genetic algorithm (GA) based approaches are employed to solve the problem by emphasizing the strategic and tactical level de-cisions. Furthermore, a design of experiment setting is constructed based on a practical case study to obtain remarkable insights regarding the HFSP problem with the lwt constraint. By doing so, the combinations of the lot streaming methodologies along with the sublot assignment methods are compared to each other through the employed algorithms. The computational results demonstrate that regardless of the employed crossover operator, the random-based sublot assignment with consistent sublots methodology leads to higher-quality solutions by up to 60 % compared to the combination of the production leveling-based sublot assignment with equal sublots methodology. The results of the experiments provide remarkable insights regarding the factors and their levels as well. In this manner, the number of sublots and the number of machines have a significant impact on the quality of the solutions. Therefore, it is suggested to implement an order release mechanism for the multi-stage manufacturing system with the hybrid flowshop setting.