A microstate spatial-inference model for network-traffic estimation

In an Advanced Traveler Information System (ATIS), sensors are often used to monitor and obtain traffic information on a real-time basis. Knowing that traffic sensors cover only a fraction of the road network, the authors investigate how to estimate traffic volumes on arcs that are not covered by sensors. By exploiting the spatial properties and the topology of a network, they derive a microstate model that can be used to estimate these traffic volumes. Based on entropy maximization, they present a microstate surrogate for competing techniques such as traffic assignment, and algebraic method or topological approach in estimating traffic flow. Being an entropy model, it also has advantage over these competing techniques in terms of the prerequisite information required to enable the model. Being a microstate rather than a steady-state model, it takes into account the fluctuation of traffic and it executes fast enough to allow real-time estimation of traffic flow. By covering the entire network flow this way with only a limited number of sensors, it will help in better driver routing decisions and traffic management tactics while being cognizant of today’s budgetary constraints facing operating agencies. The algorithm has been tested successfully in Little Rock, Arkansas and in a controlled experiment with a randomly generated 100-node/522-arc grid network.

[1]  Henry X. Liu,et al.  Optimal Detector Placement for Freeway Bottleneck Identification , 2008 .

[2]  Monica Gentili,et al.  Survey of Models to Locate Sensors to Estimate Traffic Flows , 2011 .

[3]  Zhong-Ren Peng,et al.  Rule-Based Forecasting of Traffic Flow for Large-Scale Road Networks , 2012 .

[4]  E. Jaynes.  On the rationale of maximum-entropy methods , 1982, Proceedings of the IEEE.

[5]  Ariel Caticha,et al.  Updating Probabilities with Data and Moments , 2007, ArXiv.

[6]  Srinivas Peeta,et al.  Identification of vehicle sensor locations for link-based network traffic applications , 2009 .

[7]  Praveen Edara,et al.  Optimizing Freeway Traffic Sensor Locations by Clustering Global-Positioning-System-Derived Speed Patterns , 2010, IEEE Transactions on Intelligent Transportation Systems.

[8]  Lei Zhang,et al.  Behavioral Foundation of Route Choice and Traffic Assignment , 2011 .

[9]  Waldo R. Tobler,et al.  Computation of the correspondence of geographical patterns , 1965 .

[10]  Amos Golan,et al.  Maximum Entropy, Likelihood and Uncertainty: A Comparison , 1998 .

[11]  Manwo Ng.  Synergistic sensor location for link flow inference without path enumeration: A node-based approach , 2012 .

[12]  Huapu Lu,et al.  A K-shortest-paths-based algorithm for stochastic traffic assignment model and comparison of computation precision with existing methods , 2005 .

[13]  Wanli Min,et al.  Real-time road traffic prediction with spatio-temporal correlations , 2011 .

[14]  Enrique F. Castillo,et al.  The Observability Problem in Traffic Network Models , 2008, Comput. Aided Civ. Infrastructure Eng..

[15]  C. E. SHANNON,et al.  A mathematical theory of communication , 1948, MOCO.

[16]  Oriol Serch,et al.  Exploring Link Covering and Node Covering Formulations of Detection Layout Problem , 2012 .

[17]  Yiannis Kamarianakis,et al.  Space-time modeling of traffic flow , 2002, Comput. Geosci..