Self-Learning Artificial Bee Colony Algorithm for Distributed Hybrid Flow Shop Scheduling with Preventive Maintenance

Due to advantages such as flexibility, higher production efficiency, and lower risk resilience, the distributed hybrid flow shop scheduling problem (DHFSP), as an extension of the flow shop scheduling problem, has received considerable attention from both academia and industry. However, the deterioration effects caused by wear and tear of machines and corresponding preventive maintenance (PM) strategies that can improve the deterioration effects are typically ignored in the vast majority of DHFSPs. As a consequence, a DHFSP with deteriorating effects and PM (DHFSP-DP) targeted at minimizing the makespan is first investigated in this study. To efficiently solve the DHFSP-DP, a self-learning artificial bee colony algorithm (SABC) is then designed with the following two key innovations: (1) An improved heuristic and an average random initialization are used to solve the sub-problems of job assignment and initial processing sequence; (2) The search performance of the artificial bee colony algorithm is enhanced by using a reinforcement learning method that dynamically selects the search operator. Finally, 30 test instances are used to evaluate the performance of SABC. Experimental results show that SABC outperforms other competing algorithms in solving the proposed DHFSP-DP.