Learning Traffic Network Embeddings for Predicting Congestion Propagation

Sun, Yuqing; Jiang, Guiyuan; Lam, Siew-Kei; He, Peilan

doi:10.1109/tits.2021.3105445

Cited by 8 publications

(4 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, a few studies have explicitly incorporated traffic patterns into deep learning models to developed pattern‐aware spatiotemporal prediction models (Di et al., 2019; Leiser & Yildirimoglu, 2021; Zheng et al., 2023). While many previous studies did not explicitly model the evolution of individual congestion events and only predicted congestion based on regular time intervals (Kumar & Raubal, 2021), some researchers have considered congestion as spatiotemporally propagating events and have attempted to perform fine‐grained congestion forecasting using graph embedding (Sun et al., 2022; Wang et al., 2023) or point process models (Zhu et al., 2022).…”

Section: Literature Review and Related Workmentioning

confidence: 99%

A deep marked graph process model for citywide traffic congestion forecasting

Zhang,

Wang,

Wang

et al. 2023

Computer aided Civil Eng

View full text Add to dashboard Cite

Forecasting citywide traffic congestion on large road networks has long been a nontrivial research problem due to the challenge of modeling complex evolution patterns of congestion in highly stochastic traffic environments. Arguing that purely data‐driven methods may not perform well for congestion forecasting, we propose a deep marked graph process model for predicting the congestion indices and the occurrence time of traffic congestion events for complex signalized road networks. Traffic congestion is considered as a nonrigorous spatiotemporal extreme event. We extend the traditional point process model by integrating a specially designed spatiotemporal graph convolutional network. This hybrid strategy takes advantage of the simple form of the point process model as well as the ability of graph neural networks to emulate the evolution of congestion. Experiments on real‐world congestion data sets show that the proposed method outperforms state‐of‐the‐art baseline methods, yielding satisfactory prediction results on a large signalized road network with superior computational efficiency.

show abstract

Section: Literature Review and Related Workmentioning

confidence: 99%

A deep marked graph process model for citywide traffic congestion forecasting

Zhang,

Wang,

Wang

et al. 2023

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…Urban traffic congestion has become a serious problem in recent years [1][2][3], which has resulted in extra carbon emissions, energy consumption [4], and travel time [5]. In 2019, urban Americans experienced an excess of 36 million tons of greenhouse gas emissions and 3.5 billion gallons of fuel consumption because of traffic congestion.…”

Section: Introductionmentioning

confidence: 99%

Offset Optimization Model for Signalized Intersections Considering the Optimal Location Planning of Bus Stops

Luo

Shi

2023

Systems

View full text Add to dashboard Cite

Existing offset optimization methods for signalized intersections are mainly focused on regular traffic flow, which cannot accommodate cars and public transit (e.g., Bus Rapid Transit (BRT)) simultaneously. This study proposes a delay prediction model to formulate the signal delay of BRT at intersections. The relation among the green wave bandwidth, signal timing plans, speed of the BRT vehicles, distance between the intersections, and the offset is also modeled. A combinatorial optimization model is then established, which takes the location planning of BRT stops and the offset of intersections at both directions along the artery as the decision variables. The proposed model is programmed with Mathematical Programming Language (AMPL) and solved efficiently by the Gurobi solver. The proposed optimization method is compared with seven different methods. The results show that the average BRT travel time is reduced by at least 19% and the green wave bandwidth is increased by around 30.2%. The importance of considering location planning of BRT stops when optimizing the offset is thereby verified.

show abstract

“…We will present two approaches to achieve accurate CEP. Firstly, we introduce a congestion propagation prediction model called AE-LPGT [38], which uses a deep learning-based embedding method to characterize and predict congestion propagation on network structure. AE-LPGT considers the asymmetry of propagation, local proximity with neighborhoods, and global propagation tendency over the entire network.…”

Section: Congestion Evolution Predictionmentioning

confidence: 99%

“…Extensive experiments conducted on real traffic data in Singapore, demonstrate that our method significantly outperforms the state-of-the-art baselines with prediction accuracy from 86.7% to 88.7%, and F1-score from 0.821 to 0.851. This work is published in [38].…”

mentioning

confidence: 99%

Congestion-evolution aware road traffic prediction

Sun¹

Self Cite

View full text Add to dashboard Cite

From 23-Jul-2016 until now, I have grown up a lot. I am very lucky on the way to pursue the doctorate, with the help of my supervisors, my friends, and my family. First, I would love to express my sincere gratitude to my supervisor Prof. Lam Siew-Kei, for his unlimited patience and experienced instructions. The countless discussions with him help me to understand that how to start research and overcome the challenges. The quality of my works improved gradually over these years because of his insights and advice. Many times, Prof. Lam helped me revise the paper and respond to the reviewers, despite his busy schedule. With the extensive impact of COVID-19, Prof. Lam always provides generous support whenever I meet difficulties. I cannot appreciate more about all of these givings. Also, I would like to express my special appreciation and thanks to my co-supervisor, Prof. Thambipillai Srikanthan, for his support of my Ph.D. study. With his long experience and deep expertise in the transpiration field, Prof. Srikanthan helps me to build a clear, elegant, and logical big picture of the Ph.D. project. During the regular meeting throughout my Ph.D. study, he has provided innumerable comments and suggestions that are inspiring and invaluable.I would also like to thank Dr. Jiang Guiyuan for his significant role in my candidature.He is an excellent researcher in the computer science area, with rich knowledge and experience. He patiently helped me be equipped for research in the first two years. Every time I raised a problem, he always provided exact explanations for both mathematical concepts and techniques. Dr. Jiang's instructions are the best example of "teaching one to fish is better than to give him fish," which will benefit my following research life. In the last three years of my candidature, Dr. Jiang countlessly helped me improve my works, and finally, he helped me obtain positive feedback with encouragement. I appreciate his kindness. Thanks to my dearest parents, they are the most powerful backup of my life, I vi can always be their little girl when I meet frustrations outside. Also, I want to extend my thanks to my boyfriend, Dr. Fang Zheng, who was also a Ph.D. student at SCSE, NTU.Although pursuing a Ph.D. is challenging and stressful, with his company and empathy, I could be energetic, optimistic, and hopeful.

show abstract

Learning Traffic Network Embeddings for Predicting Congestion Propagation

Cited by 8 publications

References 48 publications

A deep marked graph process model for citywide traffic congestion forecasting

A deep marked graph process model for citywide traffic congestion forecasting

Offset Optimization Model for Signalized Intersections Considering the Optimal Location Planning of Bus Stops

Congestion-evolution aware road traffic prediction

Contact Info

Product

Resources

About