A novel residual graph convolution deep learning model for short-term network-based traffic forecasting

ABSTRACT

Short-term traffic forecasting on large street networks is significant in transportation and urban management, such as real-time route guidance and congestion alleviation. Nevertheless, it is very challenging to obtain high prediction accuracy with reasonable computational cost due to the complex spatial dependency on the traffic network and the time-varying traffic patterns. To address these issues, this paper develops a residual graph convolution long short-term memory (RGC-LSTM) model for spatial-temporal data forecasting considering the network topology. This model integrates a new graph convolution operator for spatial modelling on networks and a residual LSTM structure for temporal modelling considering multiple periodicities. The proposed model has few parameters, low computational complexity, and a fast convergence rate. The framework is evaluated on both the 10-min traffic speed data from Shanghai, China and the 5-min Caltrans Performance Measurement System (PeMS) traffic flow data. Experiments show the advantages of the proposed approach over various state-of-the-art baselines, as well as consistent performance across different datasets.

KEYWORDS:

• Short-term traffic forecasting
• spatial-temporal dependency
• network topology
• graph convolution
• residual long short-term memory

Acknowledgments

The authors would like to thank the editors Dr. May Yuan and Prof. Robert Weibel, and the anonymous referees for their constructive comments to improve the quality of the article. They are also grateful to Dr. James Haworth for many valuable discussions about the paper and related work.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data and codes availability statement

The data and codes that support the findings of this study are available in ‘figshare.com’ with the identifier ‘10.6084/m9.figshare.10732871’. The two datasets are public data. The website and the citation to the first dataset (SHSpeed) are provided in the ‘SHSpeed.txt’ file. The second dataset can be downloaded from the Caltrans Performance Measurement System (PeMS). Readers interested in the second dataset can apply for an account to access the data from the PeMS directly.

Additional information

Funding

This work is part of the Consumer Data Research Centre (CDRC) project supported by the UK Economic and Social Research Council (ES/L011840/1). The first author’s PhD research is jointly funded by China Scholarship Council (Grant No. 201603170309) and the Dean’s Prize from University College London.

Notes on contributors

Yang Zhang

Yang Zhang is a Ph.D. student with SpaceTimeLab for Big Data Analytics, University CollegeLondon. Her research interests include spatio-temporal data mining, deep learning, urban computing and geo-computation with applications in transport, crime and social media.

Tao Cheng

Tao Cheng is a professor in Geoinformatics at Department of Civil, Environmental and GeomaticEngineering, University College London. She is the Founder and Director of SpaceTimeLab for BigData Analytics. Her research interests include network complexity, Geocomputation, space-timeanalytics and Big data mining (modelling, prediction, clustering, visualisation and simulation) withapplications in transport, crime, health, social media, and natural hazards.

Yibin Ren

Yibin Ren is a postdoctoral fellow at the Institute of Oceanography, Chinese Academy of Science.He visited the SpaceTimeLab for Big Data Analytics, University College London during Oct. 2017-Oct. 2018 as a jointly PhD student. His research interests include spatio-temporal data modellingand AI-aided information mining for ocean big data.

Kun Xie

Kun Xie is currently an Assistant Professor in the Department of Civil and Environmental Engineering at Old Dominion University (ODU). He received the B.S. and M.S. degrees in transportation engineering from Tongji University and the Ph.D. degree in transportation planning and engineering from New York University. His research concentrates on the use of data-driven approaches and emerging technologies to enhance the safety, efficiency and resiliency of transportation systems. Prior to joining ODU, he worked as a Lecturer (equivalent to tenure-track Assistant Professor in US) in the Department of Civil and Natural Resources Engineering at University of Canterbury and as a Postdoctoral Associate in the Center for Urban Science and Progress (CUSP) and Department of Urban and Civil Engineering at New York University. He has published over 70 refereed papers in scholarly journals and conference proceedings and is recognized by research awards such as IEEE ITS Best Dissertation Award, CUTC Milton Pikarsky Memorial Award and Transportation Research Board (TRB) Best Paper Award. He has served as a guest/area editor, a session chair and a reviewer for over 30 scientific journals and international conferences.