Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 01005 | |
Number of page(s) | 7 | |
Section | Traffic Prediction and Analysis | |
DOI | https://doi.org/10.1051/itmconf/20257001005 | |
Published online | 23 January 2025 |
Research on the Application of Reinforcement Learning in Traffic Flow Prediction
College of Intelligent and Computing, Tianjin University, Tianjin, 300354, China
Corresponding author: carrothuyq@tju.edu.cn
As global urbanization accelerates, urban traffic issues are becoming increasingly severe. Traffic flow prediction (TFP), as a key technology in intelligent transportation systems, aims to provide decision support and optimization plans by analyzing vehicle flow, speed, and density in road networks. However, traditional statistical models and prediction methods based on historical data exhibit many limitations when dealing with complex, dynamic, and nonlinear traffic flow data. The purpose of this paper is to discuss how Reinforcement Learning (RL) can be applied to TFP. RL optimizes strategies through interactions between agents and the environment to maximize cumulative rewards. High-dimensional state spaces and nonlinear problems can be handled with strong adaptability and strong adaptability. The paper provides a detailed review of the latest developments in Deep RL in the field of TFP, including the application of Q-learning and its variants in traffic signal control. Additionally, the article discusses the application of RL-based Long Short-Term Memory Networks, Graph Convolutional Networks (GCN), and Dynamic GCN in TFP. Although RL has achieved significant results in the field of TFP, its application still faces challenges such as data complexity, dynamics, and high computational resource consumption. The paper suggests that future research directions should include expanding abnormal data, improving model efficiency and scalability, and extending application scenarios to further advance intelligent transportation systems.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.