An improved multi-objective whale optimization algorithm for the hybrid flow shop scheduling problem considering device dynamic reconfiguration processes

Manufacturing industries frequently encounter production scheduling problems containing device dynamic reconfiguration processes (DRP). DRP refers to dynamic device adjustments (such as replacement of tools), leading to changes in the devices' actual processing time. It has a severe impact on the production schedule. Nevertheless, there is scarcely research upon hybrid flow shop scheduling problem (HFSP) with DRP. Besides, it is necessary to consider multiple conflict objectives in the HFSP. Thus, the multi-objective HFSP with DRP (MOHFSP-DRP) is significant in both theoretical research and application. This paper first proposes a multiobjective mathematical model (MOHFSP-DRP) that simultaneously considers the DRP and devices' adjustable processing modes. The bi-objective of this model is to minimize both the makespan and the whole device's energy consumption. This study then proposes an improved multi-objective whale optimization algorithm (IMOWOA) to solve the MOHFSP-DRP and obtain the Pareto-based optimal solution set. After that, to verify the proposed method's effectiveness, numerical experiments are implemented based on the real-world cases in a Chinese company's digital hot-rolling workshop. Results denote that the presented IMOWOA is superior to SPEA2 and NSGA-II. Finally, the MOHFSP-DRP model and IMOWOA are applied to a real-world hot-rolling shop successfully. The real-world cases verify the proposed IMOWOA can tackle the presented MOHFSP-DRP very well.

Automated summary

The study proposed a new multi-objective mathematical model (MOHFSP-DRP) to address the hybrid flow shop scheduling problem with DRP. An improved multi-objective whale optimization algorithm (IMOWOA) was introduced to obtain the optimal solution set. Experimental results demonstrated the superiority of IMOWOA in tackling the MOHFSP-DRP.