Load Balancing of Two-Sided Assembly Line Based on Deep Reinforcement Learning

In the complex and ever-changing manufacturing environment, maintaining long-term steady efficient work of assembly line is ultimate goal pursued by relevant enterprises, foundation which a balanced load. Therefore, this paper carries out research on two-sided balance problem (TALBP) for load balancing. At first, mathematical programming model established with objectives optimizing efficiency, smoothness index, completion time index (TAL). Secondly, deep reinforcement learning algorithm combining distributed proximal policy optimization (DPPO) convolutional neural network (CNN) proposed. Based agent structure assisted marker layer, task assignment states decisions selecting tasks are defined. Task logic reward function designed according to guide selection assignment. Finally, performance proposed verified benchmark problem.