Health Monitoring of a Truss Bridge using Adaptive Identification

Integration of structural analysis, system identification, and sensor networks provide health monitoring capabilities that benefit many aspects of infrastructure management. This work presents an adaptive identification method based on Lyapunov methods for a truss structure. Lyapunov methods in the identification provide guaranteed convergence for the parameter estimation. Identification is carried out on a simulation model based on FEA methods. The FEA model offers realistic results from a truss structure and allows verification of the identification method for a rolling load case. A rolling load scenario provides use of realistic loading of a structure so that identification methods may be extended for field use. Estimation results show that convergence of health parameters is suitable though the use of the adaptive estimation. Also, results of simulations show that the adaptive estimation methods are able to track changes over time to provide monitoring of a degrading structure.