Real-Time Freeway Network Traffic Surveillance: Large-Scale Field-Testing Results in Southern Italy

Abstract

Abstract-This paper reports on some large-scale field-testing results of a real-time freeway network traffic surveillance tool that has recently been developed to enable a number of real-time traffic surveillance tasks. This paper first introduces the related network traffic flow model and the approaches employed to traffic state estimation, traffic state prediction, and incident alarm. The field testing of the tool for these surveillance tasks in the A3 freeway of 100 km between Naples and Salerno in southern Italy is then reported in some detail. The results obtained are quite satisfactory and promising for further future implementations of the tool.

Prediction of traveller information and route choice based on real-time estimated traffic state

“…Recently, many research studies have focused on traffic state estimation problem (Wang et al 2011;Munoz et al 2003 ;Munoz et al 2006;Tampere and Immers 2007;Sun, Munoz and Horowitz 2004, Ngoduy 2008, Ngoduy 2011a. Of particular relevance to the present paper is the work of Wang and Papageorgiou (2005).…”

“…The EKF is considered as a de facto estimation algorithm for estimating state of a non-linear dynamic system and traffic flow models including CTM are nonlinear in nature. EKF has been consistently used for estimation of traffic state (Wang et al 2011;Messmer 2008, Wang andPapageorgiou 2005;Meier and Wehlan 2001;Cremer 1991). Other estimation algorithms for nonlinear systems such as particle filters and unscented Kalman filter are computationally expansive algorithms when compared with EKF.…”

“…The proposed model was designed and applied for a stretch of freeway with onramps and off-ramps. Wang et al (2011) applied this methodology for traffic state estimation in a freeway network of 100 km in Italy. Ngoduy (2008) proposed a framework that utilizes a particle filtering algorithm with a second-order traffic flow model to estimate traffic for a section of freeway; and in Ngoduy (2008;2011a) utilized an unscented Kalman filter algorithm with a macroscopic traffic flow model for freeway traffic state estimation.…”

Echtzeit-Verkehrszustandsschätzung und -Störfallerkennung auf Schnellstraßen mittels des Erweiterterten Kalman-Filters

“…Frühe Anwendungen der Verkehrsschätzung [10][11][12][13][14][15] bezogen sich hauptsächlich auf kurze Schnellstraßenabschnitte mit engmaschiger Messstellenanordnung unter Verwendung einfacher mathematischer Modelle. Spätere Arbeiten [2][3][4][5][6][7][16][17][18][19][20][21][22][23][24][25][26] verwenden umfassende Modelle und haben deshalb den Weg für die Berücksichtigung langer Strecken bzw. ausgedehnter Netze mit geringer Messpunktabdeckung geebnet [7].…”

“…Spätere Arbeiten [2][3][4][5][6][7][16][17][18][19][20][21][22][23][24][25][26] verwenden umfassende Modelle und haben deshalb den Weg für die Berücksichtigung langer Strecken bzw. ausgedehnter Netze mit geringer Messpunktabdeckung geebnet [7]. Ein generischer Entwurfsansatz für einen Schätzalgorithmus wurde in [2] vorgeschlagen und mit Realdaten [3,4,25,26] bzw.…”