Scholarlink

<p>This work deals with damage diagnosis in multistory building structures using sampled acceleration measurements. Structural damage is represented as a perturbation in the dynamic response and modeled through a Bouc–Wen hysteresis law embedded in a shear-building model. On this basis, a continuous–discrete unknown input observer is constructed that reconstructs both the state vector and a story-wise damage indicator in real time, although only discrete output samples are available. A Lyapunov analysis is used to derive explicit upper bounds on the admissible sampling interval, ensuring asymptotic convergence of the estimation error. The performance of the observer is evaluated on a five-story prototype subjected to seismic excitation. The simulations show accurate reconstruction of displacements and velocities and correct identification of the damaged stories from the sampled accelerations. In the reported scenarios, the mean square error of the proposed continuous–discrete observer is, on average, about 400 times smaller than that of a continuous-time unknown input observer used as benchmark under the same sampling conditions. These results highlight the potential of continuous–discrete unknown input observers for damage localization in building-like structures when only sampled acceleration measurements are available.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Unknown input observer for damage diagnosis under sampled outputs: application to a building-like structure

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Jesús Morales-Valdez,
  • Felipe Ramírez-Rasgado,
  • Omar Hernández-González,
  • María-Eusebia Guerrero-Sánchez,
  • Guillermo Valencia-Palomo,
  • Carlos-Manuel Astorga-Zaragoza

摘要

This work deals with damage diagnosis in multistory building structures using sampled acceleration measurements. Structural damage is represented as a perturbation in the dynamic response and modeled through a Bouc–Wen hysteresis law embedded in a shear-building model. On this basis, a continuous–discrete unknown input observer is constructed that reconstructs both the state vector and a story-wise damage indicator in real time, although only discrete output samples are available. A Lyapunov analysis is used to derive explicit upper bounds on the admissible sampling interval, ensuring asymptotic convergence of the estimation error. The performance of the observer is evaluated on a five-story prototype subjected to seismic excitation. The simulations show accurate reconstruction of displacements and velocities and correct identification of the damaged stories from the sampled accelerations. In the reported scenarios, the mean square error of the proposed continuous–discrete observer is, on average, about 400 times smaller than that of a continuous-time unknown input observer used as benchmark under the same sampling conditions. These results highlight the potential of continuous–discrete unknown input observers for damage localization in building-like structures when only sampled acceleration measurements are available.