Multi-resource constrained flexible job shop scheduling problem with fixture-pallet combinatorial optimisation

There is a lack of research on the flexible job shop scheduling problem (FJSP) considering limited fixture-pallet resources in multi-product mixed manufacturing workshops. However, field research in a leading engine manufacturer in China has revealed that fixture-pallet resources are a key factor limiting capacity breakthroughs although they play an auxiliary role in the production process. Thus, in this paper, we propose a methodology for the multi-resource constrained flexible job shop scheduling problem with fixture-pallet combinatorial optimisation. First, a mixed integer programming model with machine-fixture-pallet constraints is constructed aiming to minimize makespan. Then, a novel genetic algorithm integrated with feasibility correction strategy and self-learning variable neighbourhood search (VNS) is proposed to address the complicated scheduling problem, where the feasibility correction strategy is designed to solve potential conflict between machine selection and fixture selection chromosomes and self-learning VNS is executed to further improve the optimisation capability. Moreover, the effectiveness and efficiency of proposed algorithm are demonstrated by computational experiments with real data from cooperated engine manufacturing plant, which would provide convincing support for real production scheduling under complex scenarios.