Simultaneous Input and Parameter Estimation of Hysteretic Structural Systems Using Quasi-Monte Carlo-Simulation-Based Minimum Variance Unbiased Estimator

Abstract

This work proposes an efficient identification scheme for simultaneous input and parameter estimation of the hysteretic systems. For this purpose, a quasi-Monte Carlo (QMC)-simulation-based approach is adopted, where the sigma point generation scheme is coupled with a minimum variance unbiased estimator. In this process, additional bounds and constraints on the parameters are introduced to control the stability and convergence. The accuracy of the proposed algorithm is validated using a synthetic experiment on a frame whose nonlinear behavior is characterized by the Bouc–Wen–Baber–Noori model, that is, with degradation and pinching. Once the proposed algorithm is validated, its performance is further demonstrated using the shake table test of a full-scale bridge pier. The identified parameters in this case are utilized for damage quantification using a modified Park and Ang damage index. Overall, this study shows the robustness of the proposed algorithm for combined input estimation and condition assessment of inelastic reinforced concrete structures with a significant level of accuracy.