Abstract
This study is intended to investigate feasibility of bridge health monitoring utilizing dynamic parameters and a linear system parameter of a time series model identified from traffic-induced vibration data of bridge through a moving vehicle experiment on a real continuous steel-truss bridge. In addition to the natural frequency and damping constant of the bridge, this study adopts Damage Indicator (DI) which consists of autoregressive (AR) coefficients as a parameter to detect possible anomaly of the bridge. First, this study assesses the changes in frequency, damping constant, mode shape and DI. Second, Mahalano-bis-Taguchi System (MTS) is used to judge potential changes in frequency, damping constant and DI. DI can be estimated stably and its change due to damage is relatively large. Mahalanobis-Taguchi System (MTS) is successfully applied to make a decision on not only bridge condition but damage location. Observation demonstrates feasibility of structural diagnosis of bridges from the identified system parameters. However, the sensor location is still an important factor in identifying damage location which needs further investigations.
