Journal
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
Volume 23, Issue 9, Pages 16185-16196Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TITS.2022.3148358
Keywords
Roads; Predictive models; Data models; Recurrent neural networks; Generators; Computer architecture; Deep learning; Short-term link speed prediction; signalized urban networks; Wasserstein generative adversarial network
Categories
Funding
- National Key Research and Development Program of China [2020YFB2104001]
- National Natural Science Foundation of China [U1811463, 52072343]
- Zhejiang Provincial Natural Science Foundation (ZJNSF) [LY20E080023]
- iTensor Project by Richterska Stiftelsen [2019-00498]
- iHorse Project by KTH Digital Futures
Ask authors/readers for more resources
This paper proposes a short-term traffic speed prediction approach, called PL-WGAN, which uses Wasserstein Generative Adversarial Nets (WGAN) for data-driven traffic modeling. The proposed method combines generative and discriminative neural networks and captures spatial-temporal relations by stacking multiple neural networks. The results of the experiment demonstrate that this approach provides a scalable and effective traffic prediction solution for urban road networks.
Road link speed is often employed as an essential measure of traffic state in the operation of an urban traffic network. Not only real-time traffic demand but also signal timings and other local planning factors are major influential factors. This paper proposes a short-term traffic speed prediction approach, called PL-WGAN, for urban road networks, which is considered an important part of a novel parallel learning framework for traffic control and operation. The proposed method applies Wasserstein Generative Adversarial Nets (WGAN) for robust data-driven traffic modeling using a combination of generative neural network and discriminative neural network. The generative neural network models the road link features of the adjacent intersections and the control parameters of intersections using a hybrid graph block. In addition, the spatial-temporal relations are captured by stacking a graph convolutional network (GCN), a recurrent neural network (RNN), and an attention mechanism. A comprehensive computational experiment was carried out including comparing model prediction and computational performances with several state-of-the-art deep learning models. The proposed approach has been implemented and applied for predicting short-term link traffic speed in a large-scale urban road network in Hangzhou, China. The results suggest that it provides a scalable and effective traffic prediction solution for urban road networks.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available