Method for Vehicle Identification and Classification for Bridge Response Monitoring
Loading-induced deterioration of highway bridges, e.g. fatigue, is a well known phenomenon that results in reduced residual strength and service life. To estimate the residual strength and remaining life, it is essential to determine the real vehicle loading, i.e. count, spatial location, weight and velocity, on the bridge. We have developed an algorithm that takes acceleration response data (time histories) of the bridge under traffic load and gives the automatic vehicle count (N). Vehicles of interest are single unit trucks and tractor semi-trailer units as they are considered to cause maximum damage to bridge structures. The Automatic Vehicle Detection algorithm (AVD) models vibration signature of trucks on the basis of their physical parameters, i.e. gross weight, velocity and length. Two tuning parameters of the AVD [Threshold of detection (THD) & Time of solitary existence (TSE)] are varied to match detection results from the video traffic analysis. It is shown that with THD = 0.20 g and TSE = 1.86 sec, the average detections by algorithm and video traffic are off by very small value of Δ= 0.8. Thus, these values can be used to detect trucks amongst other vehicles in the acceleration time history. The resulting vehicle count/weight statistics associated with these vehicles will be available to subsequent stress and fatigue analyses on the structure.
KeywordsVideo Data Vibration Signature Acceleration Response Time Headway Accelerometer Data
Unable to display preview. Download preview PDF.
- 1.Bureau of Transportation Statistics, Condition of U.S. Highway Bridges: (1990–2008) http://www.bts.gov/current_topics/2009_03_18_bridge_data/html/bridges_us.html
- 2.U.S. Department of Transportation, Comprehensive Truck Size and Weight (CTS&W) Study, http://www.fhwa.dot.gov/reports/tswstudy/Vol1‐Summary.pdf (pp- 15, Table 5).