A Reinforcement Learning Driven Iterated Greedy Algorithm for Energy-Efficiency Flexible Job-Shop Scheduling Problem

The escalating environmental challenges have sparked significant interest in energy-efficient scheduling as a potent strategy for realizing sustainable development and fostering green manufacturing. This approach is characterized by its dual focus on economic efficiency and energy conservation. This paper tackles the energy-efficient scheduling of the flexible job-shop scheduling problem (EFJSP) with the dual objective of minimizing both makespan and total energy consumption. The mixed-integer linear programming (MILP) model of EFJSP is devised. A reinforcement learning driven iterated greedy algorithm (RLIGA) is proposed to solve the EFJSP. Furthermore, upon a thorough analysis of the problem's characteristics, two optimization strategies-a energy-saving strategy and an acceleration strategy-are developed to enhance the solution further. Extensive benchmark tests have substantiated the superior efficiency and significance of the RLIGA over state-of-the-art algorithms in addressing the EFJSP.