A probabilistic optimization model for allocating freeway sensors

Abstract This paper proposes a sensor location model to identify a sensor configuration that minimizes overall freeway performance monitoring errors while considering the consequences of probabilistic sensor failures. To date, existing sensor location models for freeway monitoring inherently assume that either deployed sensors never fail or the consequences of sensor failure are trivial matters. However, history has revealed that neither assumption is realistic, suggesting that ignoring failures in sensor allocation models may actually produce a significantly suboptimal configuration in the real world. Our work addresses this dilemma by developing a probabilistic optimization model that will minimize the error expectation by examining all possible failure scenarios, each with an occurrence probability. To ensure the scenario completeness and uniqueness, a sensor failure scenario is represented by using a binary string with 1 indicating an operational sensor at a given site and 0 for sensor failure or no sensor deployed. When applied to a case study network, it is shown that an optimal configuration that considers sensor failure is significantly different from an optimal configuration that ignores sensor failure, revealing that sensor failures pose non-trivial consequences on performance monitoring accuracy.

[1]  Pamela Murray-Tuite,et al.  Vehicular network sensor placement optimization under uncertainty , 2013 .

[2]  Alexandre M. Bayen,et al.  Optimal Sensor Placement for Freeway Travel Time Estimation , 2009 .

[3]  Monica Gentili,et al.  Locating sensors on traffic networks: Models, challenges and research opportunities , 2012 .

[4]  Giuseppe Confessore,et al.  A Network Based Model for Traffic Sensor Location with Implications on O/D Matrix Estimates , 2001, Transp. Sci..

[5]  Hani S. Mahmassani,et al.  Two-Stage Stochastic Model for Sensor Location Problem in a Large-Scale Network , 2008 .

[6]  Weimin Huang,et al.  Effect of Sensor Spacing on Performance Measure Calculations , 2006 .

[7]  K. Ozbay,et al.  A Clustering Based Methodology for Determining the Optimal Roadway Configuration of Detectors for Travel Time Estimation , 2006, 2006 IEEE Intelligent Transportation Systems Conference.

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

[9]  Hai Yang,et al.  Optimal traffic counting locations for origin–destination matrix estimation , 1998 .

[10]  Rajan Batta,et al.  Optimal placement of omnidirectional sensors in a transportation network for effective emergency response and crash characterization , 2014 .

[11]  Chao Yang,et al.  Models and algorithms for the screen line-based traffic-counting location problems , 2006, Comput. Oper. Res..

[12]  George F. List,et al.  An Information-Theoretic Sensor Location Model for Traffic Origin-Destination Demand Estimation Applications , 2010, Transp. Sci..

[13]  Hani S. Mahmassani,et al.  Structural analysis of near-optimal sensor locations for a stochastic large-scale network , 2011 .

[14]  Shou-Ren Hu,et al.  A generalized sensor location model for the estimation of network origin–destination matrices , 2014 .

[15]  Kaan Ozbay,et al.  Clustering-Based Methodology for Determining Optimal Roadway Configuration of Detectors for Travel Time Estimation , 2007 .

[16]  Hani S. Mahmassani,et al.  Sensor Coverage and Location for Real-Time Traffic Prediction in Large-Scale Networks , 2007 .

[17]  J. Y. Yen,et al.  Finding the K Shortest Loopless Paths in a Network , 2007 .

[18]  Xiaoning Zhang,et al.  The optimal cordon-based network congestion pricing problem , 2004 .

[19]  Pravin Varaiya,et al.  Probe Vehicle Runs or Loop Detectors? , 2007 .

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

[21]  Kaan Ozbay,et al.  South Jersey Real-Time Motorist Information System: Technology and Practice , 2004 .

[22]  Gilbert Laporte,et al.  Heuristics for the location of inspection stations on a network , 2000 .

[23]  Shengxue He.  A graphical approach to identify sensor locations for link flow inference , 2013 .

[24]  Adam Danczyk,et al.  A mixed-integer linear program for optimizing sensor locations along freeway corridors , 2011 .

[25]  Fulvio Simonelli,et al.  A network sensor location procedure accounting for o–d matrix estimate variability , 2012 .

[26]  Yanfeng Ouyang,et al.  Reliable sensor deployment for network traffic surveillance , 2011 .

[27]  Alexander Skabardonis,et al.  Performance Evaluation of Travel Time Methods for Real Time Traffic Applications , 2007 .

[28]  Henry X. Liu,et al.  Optimal Sensor Locations for Freeway Bottleneck Identification , 2009, Comput. Aided Civ. Infrastructure Eng..

[29]  Lawrence A. Klein,et al.  Traffic Detector Handbook: Third Edition - Volume I , 2006 .

[30]  David Eppstein,et al.  Finding the k Shortest Paths , 1999, SIAM J. Comput..

[31]  M. Pollack.  Letter to the Editor—The kth Best Route Through a Network , 1961 .

[32]  Hyoshin Park,et al.  Optimal number and location of Bluetooth sensors considering stochastic travel time prediction , 2015 .

[33]  Yafeng Yin,et al.  Sensor location problems in path-differentiated congestion pricing , 2015 .

[34]  Jitamitra Desai,et al.  A discrete optimization approach for locating Automatic Vehicle Identification readers for the provision of roadway travel times , 2006 .