ABSTRACT

A health assessment technique for infrastructures is presented following earthquakes using uncertainty-filled information. A system identification (SI)-based concept is used to identify the stiffness of structural elements following earthquakes. The excitation information is not required, and responses need not be measured at all dynamic degrees of freedom. To consider uncertainty in representing a structure and dynamic responses, a significantly modified unscented Kalman filter (UKF)-based approach is used. To implement UKF, a two-stage SI approach is proposed. It is verified using three-dimensional truss-frame systems. The procedure is robust and not sensitive to the variances of initial stiffness and measurement noise, the two most uncertainty-prone parameters.

摘要

地震后使用充满不确定性的信息，提出了一种基础设施健康评估技术。基于系统识别 (SI) 的概念用于识别地震后结构元件的刚度。不需要激励信息，也不需要在所有动态自由度上测量响应。为了考虑表示结构和动态响应的不确定性，使用了基于显着改进的无迹卡尔曼滤波器 (UKF) 的方法。为了实施 UKF，提出了一种两阶段的 SI 方法。使用三维桁架框架系统对其进行验证。该过程是稳健的，并且对初始刚度和测量噪声的变化不敏感，这两个最容易不确定的参数。