Macroscopic traffic flow modeling with physics regularized Gaussian process: A new insight into machine learning applications in transportation

Yuan, Yun; Zhang, Zhao; Yang, Xianfeng; Zhe, Shandian

doi:10.1016/j.trb.2021.02.007

Cited by 83 publications

(25 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Third, how to integrate physical knowledge into the learning methods deserves more investigation. Recently, the concept of physics regularized machine learning has been proposed for macroscopic traffic flow modeling (41,42), which is also a promising direction for microscopic behavior modeling. By introducing physical knowledge of analytical models like IDM and Gipps' model, the learning-based model might facilitate a fast training process, accuracy modeling, and convenient implementation.…”

Section: Conclusion and Further Researchmentioning

confidence: 99%

Learning-Based Stochastic Driving Model for Autonomous Vehicle Testing

Liu

Feng

et al. 2021

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

In the simulation-based testing and evaluation of autonomous vehicles (AVs), how background vehicles (BVs) drive directly influences the AV’s driving behavior and further affects the test results. Most existing simulation platforms use either predetermined trajectories or deterministic driving models to model BV behaviors. However, predetermined BV trajectories cannot react to AV maneuvers, and deterministic models are different from real human drivers because of the lack of stochastic components and errors. Both methods lead to unrealistic traffic scenarios. This paper presents a learning-based stochastic driving model that meets the unique needs of AV testing (i.e., interactive and human-like stochasticity). The model is built based on the long short-term memory architecture. By incorporating the concept of quantile regression into the loss function of the model, the stochastic behaviors are reproduced without prior assumption of human drivers. The model is trained with the large-scale naturalistic driving data (NDD) from the Safety Pilot Model Deployment project and compared with a stochastic intelligent driving model (IDM). Analysis of individual trajectories shows that the proposed model can reproduce more similar trajectories of human drivers than IDM. To validate the ability of the proposed model in generating a naturalistic driving environment, traffic simulation experiments are implemented. The results show that traffic flow parameters such as speed, range, and time headway distribution match closely with the NDD, which is of significant importance for AV testing and evaluation.

show abstract

Section: Conclusion and Further Researchmentioning

confidence: 99%

Learning-Based Stochastic Driving Model for Autonomous Vehicle Testing

Liu

Feng

et al. 2021

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

show abstract

“…Such a hybrid paradigm has gained increasing interests recently. Yuan et al [33] proposed to leverage a hybrid framework, physics regularized Gaussian process [34] for macroscopic traffic flow modeling and TSE. The hybrid methods using the PIDL framework [35], [36], [37] recently becomes an active field.…”

Section: Related Work Of Traffic State Estimationmentioning

confidence: 99%

A Physics-Informed Deep Learning Paradigm for Traffic State and Fundamental Diagram Estimation

Shi,

Mo,

Huang

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Therefore, a reliable traffic state estimation (TSE) method that could resolve the limitation of traffic data availability and coverage, is strongly needed for better traffic state prediction. Traffic state estimation (TSE) has the potential to solve this limitation (Yuan et al, 2020;Zhang et al, 2020). Such a method will promote the implementation of more successful traffic predictions on the freeway system, and thereby providing the transportation agencies more useful information in terms of making timely countermeasures and creating a much safer and efficient traffic environment.…”

Section: Literature Reviewmentioning

confidence: 99%

“…With the advance of data collecting, processing, and computation technologies in recent years, data-driven approaches have been widely developed and implemented for the TSE problem. Machine learning (ML) models are prevailing in capturing the stochastic characteristics of traffic flow with a massive amount of data (Duan et al, 2016;Li et al, 2013;Ni and Leonard, 2005;Polson and Sokolov, 2018;Tak et al, 2016;Tan et al, 2014;Tang et al, 2015;Wu et al, 2018;Yuan et al, 2020;Zhang et al, 2020;Zhong et al, 2004). The performance of ML models heavily relies on high-quality data due to the data-driven nature.…”

Section: Literature Reviewmentioning

confidence: 99%

A Hybrid Machine Learning Approach for Freeway Traffic Speed Estimation

Zhang

Yuan

Yang

2020

Transportation Research Record: Journal of the Transportation R

Self Cite

View full text Add to dashboard Cite

Accurate and timely estimation of freeway traffic speeds by short segments plays an important role in traffic monitoring systems. In the literature, the ability of machine learning techniques to capture the stochastic characteristics of traffic has been proved. Also, the deployment of intelligent transportation systems (ITSs) has provided enriched traffic data, which enables the adoption of a variety of machine learning methods to estimate freeway traffic speeds. However, the limitation of data quality and coverage remain a big challenge in current traffic monitoring systems. To overcome this problem, this study aims to develop a hybrid machine learning approach, by creating a new training variable based on the second-order traffic flow model, to improve the accuracy of traffic speed estimation. Grounded on a novel integrated framework, the estimation is performed using three machine learning techniques, that is, Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Artificial Neural Network (ANN). All three models are trained with the integrated dataset including the traffic flow model estimates and the iPeMS and PeMS data from the Utah Department of Transportation (DOT). Further using the PeMS data as the ground truth for model evaluation, the comparisons between the hybrid approach and pure machine learning models show that the hybrid approach can effectively capture the time-varying pattern of the traffic and help improve the estimation accuracy.

show abstract

Macroscopic traffic flow modeling with physics regularized Gaussian process: A new insight into machine learning applications in transportation

Cited by 83 publications

References 69 publications

Learning-Based Stochastic Driving Model for Autonomous Vehicle Testing

Learning-Based Stochastic Driving Model for Autonomous Vehicle Testing

A Physics-Informed Deep Learning Paradigm for Traffic State and Fundamental Diagram Estimation

A Hybrid Machine Learning Approach for Freeway Traffic Speed Estimation

Contact Info

Product

Resources

About