Structural system identification with the consideration of model validity by taking advantage of data from fiber-optic sensors

这项研究的目的是利用光纤传感器实现结构健康监测(SHM)，该监测可以考虑数值模型的有效性，从而实现多个和多变量的数据采集，例如使用同一光纤进行静态和动态应变数据采集。从实验和分析验证以及现场测量的结果中，成功地证明了多变量和多变量数据采集的可能性及其误差特征，通过对光纤传感器分布式应变数据的特征提取来实现所构建的边界条件监测算法的有效性，以及基于贝叶斯推理的模型校准方法的基本有效性。

This study was for the purpose of realizing a structural health monitoring (SHM) that could take into account the validity of numerical models by taking advantage of the fiber-optic sensor, which enable the multiple and multivariate data acquisitions, e.g., static and dynamic strain data acquisition using a same optical fiber. From results of the experimental and analytical verifications, and the field measurements, the possibility of multiple and multivariate data acquisitions and their error characteristics, the availability of the constructed boundary condition monitoring algorithm that was realized by the feature extraction of distributed strain data from the fiber-optic sensor, and the basic effectiveness of the Bayesian inference based model calibration method were successfully shown.