Reinforcement learning (RL) has recently gained significant attention as a dynamic method for controlling traffic signals based on the real-time traffic flow. However, previous studies paid little attention to the coefficients of the reward function, and there has been limited research investigating the effect of the railway information. To address these research gaps, this study investigates the RL-based traffic signal control considering the railway information. To achieve this, the SUMO (Simulation of Urban Mobility) software was used to establish an intelligent transportation system (ITS) field consisting of two signalized intersections and two railway crossings. Two RL algorithms, namely PPO and DQN, were employed to establish the RL-SOMO model for simulation experiments under given traffic conditions. Simulation results demonstrate that the proposed RL-based traffic signal control method performs significantly better compared to the fixed control method, which indicates the promising application of RL-based traffic control in real traffic flow control.
Reinforcement learning based traffic signal control considering the railway information in Japan
2023-09-24
703929 byte
Conference paper
Electronic Resource
English
Reinforcement learning-based traffic signal control
IEEE | 2024
|Deep Reinforcement Learning-Based Traffic Signal Control
ASCE | 2024
|European Patent Office | 2024
|Deep Reinforcement Learning-based Traffic Signal Control
IEEE | 2023
|Traffic signal control method based on reinforcement learning
European Patent Office | 2023
|