In-pavement wireless sensor network for vehicle classification

Vehicle classification data, especially for trucks, is of considerable use to agencies involved in almost all aspects of transportation and pavement engineering. Current technologies for classification involve expensive installation and calibration procedures. A wireless sensor network (WSN) for veh...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2011 10th International Conference on Information Processing in Sensor Networks S. 85 - 96
Hauptverfasser: Bajwa, Ravneet, Rajagopal, Ram, Varaiya, Pravin, Kavaler, Robert
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.04.2011
Schlagworte:
Accelerometers
Axle Detection
Axles
Energy Detector
Sensor Networks
Sensors
Synchronization
Traffic Monitoring
Vehicle Classification
Vehicles
Vibrations
Wireless sensor networks
ISBN:9781612848549, 1612848540
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Vehicle classification data, especially for trucks, is of considerable use to agencies involved in almost all aspects of transportation and pavement engineering. Current technologies for classification involve expensive installation and calibration procedures. A wireless sensor network (WSN) for vehicle classification based on axle count and spacing was designed, calibrated, tested, and deployed near a weigh station in Sunol, California. The WSN includes: vibration sensors which report pavement acceleration; vehicle detection sensors which report vehicle's time of arrival and departure; and an Access Point (AP) that logs the data collected from all these sensors. Both sensors are packaged for durability, occupy minimal space, have long lifetimes, and are embedded inside the pavement. The vibration sensors are capable of over-the-air software programming and are designed to be immune to sound. Vibration and classification ground truth data for 53 different trucks exiting the weigh station were collected. The vibration data collected at 512 Hz had an accuracy of 400 μg. A novel algorithm for estimating axle count and spacing has been developed. The combination of bandwidth-aware smoothing filter and peak detector that we use in this algorithm could be useful in many other applications. The algorithm successfully classified all 53 trucks.
ISBN:9781612848549
1612848540