CHARACTERIZING TRAFFIC EXCITATION FOR HIGHWAY BRIDGE STRUCTURAL HEALTH MONITORING PURPOSE

Vibration-based structural health monitoring of highway bridges relies on identification of bridge structural condition from bridge vibration signals. To capture the long-term and in-situ bridge condition evolution without interrupting normal traffic service on a bridge, ambient vibration monitoring is usually performed, where the excitation on the bridge is neither controlled nor measured. In this case, a statistical description on the excitation characteristics is required. This paper presents an innovative approach to obtain traffic information by processing the traffic flow images and hence characterize the traffic excitation on a bridge by a stochastic model. This model enables a Bayesian updating procedure upon the parameters that represent the bridge structural integrity, utilizing information obtained both from vibration measurement and from traffic flow image processing to reduce the uncertainty in the assessment of bridge structural condition.