희소 센서 데이터를 활용한 기하 비선형 구조물의 동적 응답 업데이트 및 강건성 평가

초록

Estimating the dynamic response of real-world structures is challenging because of limited measurement data and model mismatch caused by uncertainties. This study proposes a procedure for correcting the full-field displacement response of nonlinear structures using proper orthogonal decomposition-based reduced-order representations and sensor residuals. The effectiveness of the proposed method was evaluated using a large deformation cantilever beam model, focusing on varying sensor counts and noise levels. The results indicate that the correction performance significantly improves as the number of sensors increases, with a sharp reduction in error observed with 10 or more sensors, specifically for this example. Furthermore, even under a 10% noise condition, the relative error after correction remained at 10.8%, demonstrating the robustness and validity of the proposed procedure for nonlinear response estimation.

키워드