As the critical component of manufacturing systems, production scheduling aims to optimize objectives in terms of profit, efficiency, and energy consumption by reasonably determining the main factors including processing path, machine assignment, execute time and so on. Due to the large scale and strongly coupled constraints nature, as well as the real-time solving requirement in certain scenarios, it faces great challenges in solving the manufacturing scheduling problems. With the development of machine learning, Reinforcement Learning (RL) has made breakthroughs in a variety of decision-making problems. For manufacturing scheduling problems, in this paper we summarize the designs of state and action, tease out RL-based algorithm for scheduling, review the applications of RL for different types of scheduling problems, and then discuss the fusion modes of reinforcement learning and meta-heuristics. Finally, we analyze the existing problems in current research, and point out the future research direction and significant contents to promote the research and applications of RL-based scheduling optimization.

This paper addresses the Energy-Aware Distributed Hybrid Flow Shop Scheduling Problem with Multiprocessor Tasks (EADHFSPMT) by considering two objectives simultaneously, i.e., makespan and total energy consumption. It consists of three sub-problems, i.e., job assignment between factories, job sequence in each factory, and machine allocation for each job. We present a mixed inter linear programming model and propose a Novel Multi-Objective Evolutionary Algorithm based on Decomposition (NMOEA/D). We specially design a decoding scheme according to the characteristics of the EADHFSPMT. To initialize a population with certain diversity, four different rules are utilized. Moreover, a cooperative search is designed to produce new solutions based on different types of relationship between any solution and its neighbors. To enhance the quality of solutions, two local intensification operators are implemented according to the problem characteristics. In addition, a dynamic adjustment strategy for weight vectors is designed to balance the diversity and convergence, which can adaptively modify weight vectors according to the distribution of the non-dominated front. Extensive computational experiments are carried out by using a number of benchmark instances, which demonstrate the effectiveness of the above special designs. The statistical comparisons to the existing algorithms also verify the superior performances of the NMOEA/D.