Physics-Informed Deep Learning for Traffic State Estimation

Traffic state estimation (TSE), which reconstructs the traffic variables (e.g., density) on road segments using partially observed data, plays an important role on efficient traffic control and operation that intelligent transportation systems (ITS) need to provide to people. Over decades, TSE approaches bifurcate into two main categories, model-driven approaches and data-driven approaches. However, each of them has limitations: the former highly relies on existing physical traffic flow models, such as Lighthill-Whitham-Richards (LWR) models, which may only capture limited dynamics of real-world traffic, resulting in low-quality estimation, while the latter requires massive data in order to perform accurate and generalizable estimation. To mitigate the limitations, this paper introduces a physics-informed deep learning (PIDL) framework to efficiently conduct high-quality TSE with small amounts of observed data. PIDL contains both model-driven and data-driven components, making possible the integration of the strong points of both approaches while overcoming the shortcomings of either. This paper focuses on highway TSE with observed data from loop detectors, using traffic density as the traffic variables. We demonstrate the use of PIDL to solve (with data from loop detectors) two popular physical traffic flow models, i.e., Greenshields-based LWR and three-parameter-based LWR, and discover the model parameters. We then evaluate the PIDL-based highway TSE using the Next Generation SIMulation (NGSIM) dataset. The experimental results show the advantages of the PIDL-based approach in terms of estimation accuracy and data efficiency over advanced baseline TSE methods.

[1]  Zheng Wang,et al.  Physics Regularized Gaussian Processes , 2020, ArXiv.

[2]  Shandian Zhe,et al.  Macroscopic Traffic Flow Modeling with Physics Regularized Gaussian Process: A New Insight into Machine Learning Applications , 2020, Transportation Research Part B: Methodological.

[3]  George Em Karniadakis,et al.  Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations , 2020, Science.

[4]  Yinhai Wang,et al.  Missing data imputation for traffic flow based on combination of fuzzy neural network and rough set theory , 2020, J. Intell. Transp. Syst..

[5]  Wei Huang,et al.  Traffic parameters estimation for signalized intersections based on combined shockwave analysis and Bayesian Network , 2019, Transportation Research Part C: Emerging Technologies.

[6]  Zibin Zheng,et al.  Deep and Embedded Learning Approach for Traffic Flow Prediction in Urban Informatics , 2019, IEEE Transactions on Intelligent Transportation Systems.

[7]  Paris Perdikaris,et al.  Adversarial Uncertainty Quantification in Physics-Informed Neural Networks , 2018, J. Comput. Phys..

[8]  Hyochoong Bang,et al.  Introduction to Kalman Filter and Its Applications , 2018, Introduction and Implementations of the Kalman Filter.

[9]  Anuj Karpatne,et al.  Physics Guided Recurrent Neural Networks For Modeling Dynamical Systems: Application to Monitoring Water Temperature And Quality In Lakes , 2018, ArXiv.

[10]  Paris Perdikaris,et al.  Machine Learning of Space-Fractional Differential Equations , 2018, SIAM J. Sci. Comput..

[11]  Michael S. Triantafyllou,et al.  Deep learning of vortex-induced vibrations , 2018, Journal of Fluid Mechanics.

[12]  Deepthi Mary Dilip,et al.  Learning Traffic Flow Dynamics Using Random Fields , 2018, IEEE Access.

[13]  Maziar Raissi,et al.  Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations , 2018, J. Mach. Learn. Res..

[14]  Yi Zhang,et al.  A Novel Traffic Flow Data Imputation Method for Traffic State Identification and Prediction Based on Spatio-Temporal Transportation Big Data , 2018 .

[15]  George E. Karniadakis,et al.  Hidden physics models: Machine learning of nonlinear partial differential equations , 2017, J. Comput. Phys..

[16]  R. Bürger,et al.  Hybrid Stochastic Galerkin Finite Volumes for the Diffusively Corrected Lighthill-Whitham-Richards Traffic Model , 2017 .

[17]  Nagiza F. Samatova,et al.  Theory-Guided Data Science: A New Paradigm for Scientific Discovery from Data , 2016, IEEE Transactions on Knowledge and Data Engineering.

[18]  Nicholas G. Polson,et al.  Deep learning for short-term traffic flow prediction , 2016, 1604.04527.

[19]  Hwasoo Yeo,et al.  Data-Driven Imputation Method for Traffic Data in Sectional Units of Road Links , 2016, IEEE Transactions on Intelligent Transportation Systems.

[20]  R. Bürger,et al.  Efficient parameter estimation in a macroscopic traffic flow model by discrete mollification , 2015 .

[21]  Barak A. Pearlmutter,et al.  Automatic differentiation in machine learning: a survey , 2015, J. Mach. Learn. Res..

[22]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[23]  Bin Ran,et al.  Robust Missing Traffic Flow Imputation Considering Nonnegativity and Road Capacity , 2014 .

[24]  Benjamin Seibold,et al.  Comparative model accuracy of a data-fitted generalized Aw-Rascle-Zhang model , 2013, Networks Heterog. Media.

[25]  Li Li,et al.  Efficient missing data imputing for traffic flow by considering temporal and spatial dependence , 2013 .

[26]  Raimund Bürger,et al.  Regularized Nonlinear Solvers for IMEX Methods Applied to Diffusively Corrected Multispecies Kinematic Flow Models , 2013, SIAM J. Sci. Comput..

[27]  Daniel B. Work,et al.  Phase transition model of non-stationary traffic flow: Definition, properties and solution method , 2013 .

[28]  Benjamin Seibold,et al.  Data-Fitted First-Order Traffic Models and Their Second-Order Generalizations , 2013 .

[29]  Marcello Montanino,et al.  Making NGSIM Data Usable for Studies on Traffic Flow Theory , 2013 .

[30]  Alexandre M. Bayen,et al.  Arterial travel time forecast with streaming data: A hybrid approach of flow modeling and machine learning , 2012 .

[31]  A. Bayen,et al.  On sequential data assimilation for scalar macroscopic traffic flow models , 2012 .

[32]  Jorge A. Laval,et al.  Stochastic Extension of Newell's Three-Detector Method , 2012 .

[33]  Jia Li,et al.  Traffic viscosity due to speed variation: Modeling and implications , 2010, Math. Comput. Model..

[34]  Ludovic Leclercq,et al.  A mechanism to describe the formation and propagation of stop-and-go waves in congested freeway traffic , 2010, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[35]  Yoshua Bengio,et al.  Understanding the difficulty of training deep feedforward neural networks , 2010, AISTATS.

[36]  Ludovic Leclercq,et al.  From heterogeneous drivers to macroscopic patterns in congestion , 2010 .

[37]  Xuan Di,et al.  Hybrid Extended Kalman Filtering Approach for Traffic Density Estimation along Signalized Arterials , 2010 .

[38]  William E. Schiesser,et al.  Linear and nonlinear waves , 2009, Scholarpedia.

[39]  Markos Papageorgiou,et al.  Real-time freeway traffic state estimation based on extended Kalman filter: Adaptive capabilities and real data testing , 2008 .

[40]  Andreas Hegyi,et al.  An unscented Kalman filter for freeway traffic estimation , 2006 .

[41]  L. Mihaylova,et al.  A particle filter for freeway traffic estimation , 2004, 2004 43rd IEEE Conference on Decision and Control (CDC) (IEEE Cat. No.04CH37601).

[42]  M. Zhong,et al.  ESTIMATION OF MISSING TRAFFIC COUNTS USING FACTOR, GENETIC, NEURAL AND REGRESSION TECHNIQUES , 2004 .

[43]  Dirk Helbing,et al.  Reconstructing the spatio-temporal traffic dynamics from stationary detector data , 2002 .

[44]  Benjamin Coifman,et al.  Estimating travel times and vehicle trajectories on freeways using dual loop detectors , 2002 .

[45]  P. Nelson Traveling-wave solutions of the diffusively corrected kinematic-wave model , 2002 .

[46]  H. M. Zhang A NON-EQUILIBRIUM TRAFFIC MODEL DEVOID OF GAS-LIKE BEHAVIOR , 2002 .

[47]  J. Nocedal,et al.  A Limited Memory Algorithm for Bound Constrained Optimization , 1995, SIAM J. Sci. Comput..

[48]  Jeong Gyu Kang,et al.  Estimating destination-specific traffic densities on urban freeways for advanced traffic management , 1994 .

[49]  C. Daganzo THE CELL TRANSMISSION MODEL.. , 1994 .

[50]  G. F. Newell A simplified theory of kinematic waves in highway traffic, part II: Queueing at freeway bottlenecks , 1993 .

[51]  S. L. Paveri-Fontana,et al.  On Boltzmann-like treatments for traffic flow: A critical review of the basic model and an alternative proposal for dilute traffic analysis , 1975 .

[52]  P. I. Richards Shock Waves on the Highway , 1956 .

[53]  M J Lighthill,et al.  On kinematic waves II. A theory of traffic flow on long crowded roads , 1955, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences.

[54]  Masao Kuwahara,et al.  Traffic State Estimation Using Traffic Measurement from the Opposing Lane—Error Analysis Based on Fluctuation of Input Data , 2019, Intelligent Transport Systems for Everyone’s Mobility.

[55]  Alexandre M. Bayen,et al.  Traffic state estimation on highway: A comprehensive survey , 2017, Annu. Rev. Control..

[56]  Mauro Garavello,et al.  Models for vehicular traffic on networks , 2016 .

[57]  Tsubasa Takigawa,et al.  Estimating Vehicle Trajectories on a Motorway by Data Fusion of Probe and Detector Data: for the 20th ITS World Congress Tokyo 2013 , 2013 .

[58]  Henry X. Liu,et al.  A stochastic model of traffic flow: Theoretical foundations , 2011 .

[59]  Markos Papageorgiou,et al.  An adaptive freeway traffic state estimator , 2009, Autom..

[60]  Daiheng Ni,et al.  Markov Chain Monte Carlo Multiple Imputation Using Bayesian Networks for Incomplete Intelligent Transportation Systems Data , 2005 .

[61]  Alexander Skabardonis,et al.  Detecting Errors and Imputing Missing Data for Single-Loop Surveillance Systems , 2003 .

[62]  William T. Scherer,et al.  Exploring Imputation Techniques for Missing Data in Transportation Management Systems , 2003 .

[63]  Axel Klar,et al.  Derivation of Continuum Traffic Flow Models from Microscopic Follow-the-Leader Models , 2002, SIAM J. Appl. Math..

[64]  Harold J Payne,et al.  MODELS OF FREEWAY TRAFFIC AND CONTROL. , 1971 .

[65]  F H Green,et al.  TECHNIQUES FOR MEASURING OVER-ALL SPEEDS IN URBAN AREAS , 1950 .

[66]  B D Greenshields,et al.  A study of traffic capacity , 1935 .