Huang, Siyuan (Piedmont Hills High School), Peng, Xianyue (University of California, Davis), Zhang, H. Michael (University of California Davis)

Energy-Aware Bus Speed Control Via Proximal Policy Optimization: A Comparative Study under Fixed-Time, Bus-Priority, and Backpressure Signal Control Strategies

Scheduled for presentation during the Video Session "On-Demand Video Presentations" (VP-VP), Saturday, November 22, 2025, 08:00−18:00, On-Demand Platform

2025 IEEE 28th International Conference on Intelligent Transportation Systems (ITSC), November 18-21, 2025, Gold Coast, Australia

This information is tentative and subject to change. Compiled on April 2, 2026

Abstract

This study introduces a novel deep reinforcement learning approach for optimizing urban public transport. The approach guides bus speed across signalized intersections to reduce travel time and energy consumption. Treating the bus as an agent, a Proximal Policy Optimization (PPO) algorithm is employed to train the agent's policy. Three traffic signal control strategies—fixed-time signal control, bus-priority control and backpressure control—are integrated during model training. We develop a simulation framework coupling SUMO for traffic simulation with Ray RLlib for deep reinforcement learning (DRL) policy training with PPO. The DRL agent utilizes a neural network architecture that combines fully connected layers and LSTM to capture both spatial and temporal dynamics. The results demonstrate a significant reduction in bus travel time and energy consumption, with the backpressure strategy outperforming the other two strategies. The proposed method offers a promising solution to improve the efficiency and sustainability of urban transit systems.