Research on dynamic distributed flow shop dynamic scheduling based on LSTM-PPO

Aiming at the Dynamic Distributed Flow Shop Scheduling Problem (DDFSP), a scheduling solution method combining a Long short-term memory (LSTM) network with a Proximal Policy Optimization(PPO) algorithm is proposed. To solve the scheduling confusion caused by many uncertain factors in the scheduling process. By learning and analyzing a large amount of data in the dynamic distributed workshop system, the approximate optimal solution is obtained by integrating the time sequence feature extraction capability of the LSTM network and the optimization efficiency of the PPO algorithm, to realize more intelligent and flexible scheduling decision-making, improve production efficiency and process optimization level, and verified by experiment and simulation. The results show that this method can improve production efficiency and resource utilization, and has the potential for cost control.