A Knowledge-Driven Cooperative Optimization Algorithm for Multi-objective Energy-Efficient Flexible Job Scheduling with Variable Machine Speeds

In this study, a knowledge-driven multi-objective imperialist competitive algorithm (KMOICA) is employed to address the flexible job shop scheduling problem with time-based constraints, including transportation time, setup time, and variable machine speeds. Two objectives are considered simultaneously, i.e., minimization of make-span and total energy consumption. To solve the considered problem, first, a hybrid initialization method with four heuristics is designed to generate solutions with high quality and diversity. Then, a dynamic decoding strategy is designed to seek idle time across all available machines to enhance the convergence abilities. Next, to improve the exploitation abilities, a collaborative search strategy is proposed, which combines three knowledge-driven neighborhood structures and two velocity strategies. Additionally, a dynamic feedback mechanism is employed to further accelerate the convergence speed of the algorithm. Finally, through comprehensive computational comparisons and statistical analysis, the proposed algorithm demonstrates favorable performance in terms of solution quality and efficiency when compared to several presented algorithms.