Energy-efficient scheduling for rework systems based on real-time performance considering batch production and machine dynamics

Rework production systems are commonly encountered in real-world manufacturing environments. Despite its comparable significance to steady-state analysis, the transient behaviors of rework production systems with machine dynamics have not received as much attention. In this work, a rework system capturing the dynamics of geometric machines, finite-capacity buffers and batch production is investigated (i.e., geometric rework systems). To this end, the transient model along with exact performance analysis methods are firstly presented based on Markovian approach. For large-scale and multi-loop rework production systems, a novel equivalent aggregation approach is proposed for the system decomposition of the geometric rework system. On this basis, the real-time performance of the geometric rework system can be approximated with high accuracy. Moreover, a multi-objective grey wolf optimizer with weighted guidance and enhanced searching operators (MOGWO-WE) is developed, which is tested through 18 benchmarks of the formulated energy-efficient scheduling problems and outperforms the baseline algorithms with high-quality Pareto front.