A framework of cloud-edge collaborated digital twin for flexible job shop scheduling with conflict-free routing

Abstract

Synchronous scheduling of transportation resources such as automated guided vehicles (AGVs) and production resources can effectively improve the production efficiency of an automated flexible job shop, however limited transportation resources and persistent abnormal disturbances compromise the accuracy, reliability and practicability of such scheduling. In order to address the flexible job shop scheduling problem with limited transportation resources (FJSP-T) under abnormal disturbances, here we propose a cloud-edge combined digital twin (DT) flexible job shop scheduling (CE-DTFJSS) framework to realize real-time scheduling. Firstly, a time-space network based zero-one programming (TSN-ZOP) model is established to optimize makespan and energy consumption of flexible job shop scheduling with conflict-free routing. Subsequently, to achieve real-time interaction between scheduling system and physical workshop, a cloud-edge architecture is introduced to allocate computing resources for scheduling-supporting tasks such as simulation, state monitoring and distributed control in the DT system. Additionally, we design a cloud edge collaboration rescheduling strategy based on a rolling window, to update the parameters’ values in the improved time window Dijkstra nested multi-layer coding genetic algorithm we proposed, thus real-time scheduling can be achieved. The effectiveness of CE-DTFJSS are verified by a numeric simulation based on a real-world case of a digitalized automobile welding workshop. The experimental results indicate that compared with the scheduling scheme without considering AGVs routing problem, the CE-DTFJSS framework can reduce makespan and energy consumption by 8.30% and 3.73%, respectively, and real-time responses to abnormal disturbances can be rendered within 0.1 s.