FE-based bridge weigh-in-motion based on an adaptive augmented Kalman filter

Precise knowledge of the moving forces acting on bridges is essential for bridge design and maintenance. Existing studies fall short in comprehensively integrating finite element (FE) model updating and bridge weigh-in-motion (B-WIM) for accurate force identification. Therefore, this study introduces an FE-based B-WIM framework that employs an adaptive augmented Kalman filter (AAKF) to address multiple uncertainties and different vehicle configurations. The framework is composed of two essential elements: (i) updating of bridge structural parameters in the FE model utilizing Bayesian methods, and (ii) estimation of vehicle axle loads via the AAKF combining the updated FE model, axle positions, and measured/simulated bridge response data. A new adaptive noise filter based on genetic algorithm optimization is applied to provide high estimation accuracy of the load for diverse vehicle configurations and velocities. Numerical examples of a simply-supported bridge and a three-span continuous bridge are provided. The effect of the position noise level, bridge response noise level, vehicle velocity, and vehicle axle configuration on the accuracy of the identification results are comprehensively investigated. The results demonstrate the robustness and accuracy of the proposed framework under different circumstances.