Bridge Performance Evaluation via Dynamic Fingerprints and Data Fusion

Rapidly aging bridges are a serious problem plaguing the transportation industry. Bridges are subjected to accumulative damage due to environmental effects, traffic, and inappropriate management, which commonly render them structurally deficient over the course of their service life. There are drawbacks inherent to traditional visual inspection methods, including costliness and inefficiency. This paper proposes an innovative damage assessment method based on dynamic fingerprints and data fusion techniques for bridge performance evaluation. Numerical simulation is first applied to obtain dynamic fingerprints under various damage scenarios, then an ambient excitation modal test is conducted to acquire field modal data. Several data fusion techniques are then applied; Bayesian fusion, rough set theory, and Naïve Bayes classifier are combined to make full use of their advantages for convenient and efficient damage evaluation. The proposed method allows for quick detection of the existence, location, and severity of damage for the sake of highly efficient structural condition assessment. Two typical concrete continuous bridges are used to validate the effectiveness of the proposed method. One is a short-span bridge suffering minor damage, and the other is a large-span bridge suffering moderate damage. The results presented here may represent very useful information for the management and maintenance of existing bridges.