Dynamic scheduling of multi-memory process flexible job shop problem based on digital twin

Abstract

Digital twin is one of the newly-emerged enabling technologies for achieving intelligent manufacturing. Based on the physical–digital convergence, digital twin provides manufacturing systems with a new model of collaboration between the workforce and industrial processes. With the characteristics of real-time communication and data-driven enablers, the digital twin scheduling strategy requires close cooperation between workers, systems and processes. However, in the process of digitization and intelligentization, industry will need to face the challenge of supporting new technologies and worker skills development. To this end, the paper considers workers’ multi-memory process (learning and forgetting) in the flexible job shop scheduling problem (MPFJSP). Meanwhile, the dynamic scheduling strategy of the digital twin-driven MPFJSP is proposed under machine breakdowns aiming at simultaneously minimizing the makespan, total carbon emissions, total production cost and product quality stability. A virtual workshop is adopted to simulate and optimize the dynamic scheduling scheme to realize intelligent workshop scheduling. Finally, a computational experiment is carried out to verify the effectiveness and advantages of the proposed intelligent scheduling strategy.