Decentralized Loop Substructure Identification for Shear Structures with Virtual Control System

Dongyu Zhang,Zifan Niu,Jiadong Tian,Hui Li

DOI: https://doi.org/10.1002/stc.2866

2021-01-01

Abstract:Substructure identification is an effective decentralized structural health monitoring approach for large civil structures. Previously, the authors proposed a loop substructure identification method for shear structures, which utilized the dynamic equilibrium of a non-top-floor substructure to formulate a loop identification sequence (LIS), directly estimating the substructure's parameters by only using four structural acceleration responses. However, it was found that the LIS is conditional converged. To ensure the LIS's convergence, two different methods were previously investigated, but both methods have their limitations and may not guarantee satisfactory identification results under certain circumstances. In this paper, some virtual control systems (VCSs) are introduced into the loop substructure identification. The VCS is a self-balanced system, consisting of some virtual control devices and the self-balanced forces. The virtual control devices are combined with the substructure to form a controlled substructure that is used to replace the original substructure in the loop substructure identification. An in-depth error analysis of the VCS-based loop substructure identification was conducted herein, showing that the carefully designed VCS can tune the dynamics of the controlled substructure so that the convergence of the LIS can be easily achieved. Furthermore, a robust two-step design strategy is proposed to obtain the optimal VCS's parameters with uncertain structural parameters. Finally, the numerical simulation study on a 20-story shear building and the shake table tests on a five-story bench-scaled structural model were conducted to verify the effectiveness of the designed VCS to improve the convergence and identification accuracy of the LIS.