Traffic scheduling algorithm with time constraint based on proximal policy optimization (PPO) algorithm

This paper presents a traffic scheduling algorithm with time constraints based on the Proximal Policy Optimization (PPO) algorithm. The goal is to optimize vehicle routing and scheduling within a specified time window to minimize total travel time and maximize resource efficiency. PPO, a reinforcement learning method, is utilized due to its robustness in handling continuous action spaces and ensuring stable policy updates. The state space represents the current traffic conditions, vehicle locations, and available routes, while the action space corresponds to vehicle scheduling decisions, such as route selection and timing. The reward function is designed to encourage minimizing delays, reducing travel time, and meeting time constraints. By using PPO, the algorithm iteratively improves scheduling strategies by balancing exploration and exploitation, ensuring convergence to an optimal or near-optimal solution. Simulation results demonstrate the effectiveness of the proposed approach in reducing total traffic congestion and meeting scheduling constraints under dynamic traffic conditions.