Collaborative optimisation framework for multi-stage flexible assembly shop scheduling with mixed production pattern
针对家电产品多阶段柔性装配车间调度问题，提出基于分析目标级联的协同优化框架，结合改进遗传算法协调子问题，实验证明其有效降低总延迟并优于对比算法。
Diverse customisation requirements of home appliance products present significant challenges to integrated scheduling of multi-stage flexible manufacturing systems with fabrication, pre-assembly and final-assembly. Multi-stage flexible assembly shop scheduling problem considering mixed production (MFASSP-MP) requires the simultaneous coordination of subproblems with varying resource constraints, interconnected through hierarchical coupling. These subproblems include flexible job shop scheduling, flexible assembly flow shop scheduling, and unrelated parallel machine scheduling, making it difficult to achieve satisfactory solutions within finite time using integrated optimisation. To address these issues, this paper formulates a mathematical model for MFASSP-MP to minimise total tardiness, guiding decomposition and coordination. A collaborative optimisation framework based on analytical target cascading (ATC) is developed to connect subproblems, with an improved genetic algorithm (IGA) employed to correct response deviations among different levels. Furthermore, incorporating three-vector coding, ATC-IGA utilises crossover and mutation strategies to accommodate feasible domain variations and local optimisation traits, thereby enhancing global objective optimisation. To validate the effectiveness of the proposed algorithm, sufficient experiments are conducted under different scale instances. Results indicate the ATC-IGA is effective and significantly outperforms comparison algorithms. Additionally, exploring different decomposition strategies reveals the importance of balancing problem complexity to ensure convergence and diversity of algorithms for excellent performance.