A Multi-Type data driven framework for solving flexible job shop scheduling problem considering multiple production resource states

Abstract

The development of flexible manufacturing models has been propelled by Industry 4.0, making it a cornerstone of intelligent manufacturing. To address the challenges posed by frequent order changes and multiple production state disruptions in highly customized manufacturing processes. In this paper, a new framework for solving dynamic flexible job shop scheduling problem is proposed for the first time. A state constraint representation method is proposed, which can decouple the relationship between the scheduling optimization algorithm and various constraint conditions. The feasibility of the method is validated under six dynamic production states, including the shift calendar for equipment, equipment availability, equipment failures, equipment maintenance, job rework, and the insertion of jobs. Moreover, an improved Genetic Algorithm is deployed within the framework to address scheduling optimization. Compared to multiple algorithms, the proposed method is competitive in terms of optimization effectiveness and efficiency. Furthermore, the framework is deployed in a certain aerospace engine machining workshop, and the results demonstrate that the proposed framework is competitive in performing complex tasks.