Deep Reinforcement Learning-Based LSTM Model for Traffic Flow Forecasting in Internet of Vehicles
Deep Reinforcement Learning-Based LSTM Model for Traffic Flow Forecasting in Internet of Vehicles
- New search for: Chen, Zekuan
- New search for: Luo, Xiong
- New search for: Wang, Ting
- New search for: Wang, Weiping
- New search for: Zhao, Wenbing
- New search for: Deng, Zhidong
- New search for: Chen, Zekuan
- New search for: Luo, Xiong
- New search for: Wang, Ting
- New search for: Wang, Weiping
- New search for: Zhao, Wenbing
Nowadays many large cities often suffer from road congestion. Then, it is necessary to improve the working efficiency of urban road network with the help of computational intelligence technologies, within the Internet of vehicles (IoV) environment. We are committed to designing traffic flow prediction methods through the use of machine learning models. In this paper, we predict the traffic flow and traffic situation of the selected road segment. Specifically, we propose a deep reinforcement learning (DRL)-based long short-term memory (LSTM) model to predict the traffic flow. On the basis of it, we characterize the traffic situation with fuzzy comprehensive evaluation (FCE)-based model. The experimental results show that the constructed DRL-based LSTM model can accurately predict the traffic flow data. Meanwhile, the traffic situation characterization model can represent the road traffic situation well.
Document information
Deep Reinforcement Learning-Based LSTM Model for Traffic Flow Forecasting in Internet of Vehicles
Lect. Notes Electrical Eng.
Proceedings of 2021 Chinese Intelligent Automation Conference ; Chapter : 57 ; 515-523
2021-10-08
9 pages
Article/Chapter (Book)
Electronic Resource
English
Similar titles
| DOAJ | 2020
Deep Learning Mechanism for Region Based Urban Traffic Flow Forecasting
| Springer Verlag | 2023