Research on Modeling and Reinforcement Learning Scheduling Approaches for Flexible Job Shop Systems

This study focuses on the Flexible Job Shop Scheduling Problem (FJSP) and proposes an improved Proximal Policy Optimization (PPO) algorithm. First, we establish a mathematical model for flexible job shop scheduling that considers process sequences and machine resources, formulate it as a Markov Decision Process (MDP), and design the state space, action space, and reward function. Then, we enhance the traditional PPO algorithm by introducing a prioritized experience replay mechanism. Case studies demonstrate that the optimized method effectively reduces the makespan (maximum completion time). This research provides both theoretical guidance and practical significance for optimizing flexible job shop production scheduling and improving production efficiency.