VIBRATION-BASED DAMAGE IDENTIFICATION OF REINFORCED CONCRETE ARCH BRIDGES USING KALMAN–ARMA–GARCH MODEL

Shuchang Zhou,∗ Yan Jiang,∗∗ Xiaoqing Li,∗∗∗ and Qingliang Wu∗∗

Keywords

Arch bridge structure, damage identification, time series, heteroskedasticity∗ College of Electronic Information Engineering, SouthwestUniversity, Chongqing, 400715, China; email: zsc2473823148@163.com∗∗ College of Engineering and Technology, Southwest University,Chongqing, 400715, China; email: xnjtjiangyan@163.com,wuqingliang@swu.edu.cn∗∗∗ Huashe Design Group Co., Ltd., 210014, China; email: lixq@mails.cqjtu.edu.cnCorresponding author: Yan JiangRecommended by Dr. Jingz

Abstract

To ensure safe operations of bridges, it is necessary to carry out the structural damage identification and safety assessment. To this end, this paper proposes a novel damage identification method based on structural health monitoring data, which is the combination of Kalman filter, autoregressive moving average (ARMA) model and generalized autoregressive conditional heteroskedasticity (GARCH) model. Firstly, the correlation between the system characteristics and the time series model is verified through the theoretical derivation of the system vibration equation. Secondly, Kalman filtering is used to preprocess the acceleration data and reduce the noise disturbance, by which a linear recursive ARMA model can be established to identify the structural damage. Then, a nonlinear recursive GARCH model is introduced to further improve the identification accuracy. Finally, the effectiveness of the proposed method is verified using the time history data obtained from the accelerated corrosion damage dynamic test of the reinforced concrete arch. The results show that: (1) the system vibrations are correlated with the time series model, whose residual variance ratio is demonstrated to be effective in identifying structural damage; (2) in the state of loading damage and corrosion damage, the identification accuracies of Kalman–ARMA are 32.8% and 75.8%, while those of the proposed method can reach 89.1% and 85.5%, respectively and (3) GARCH model can explain the heteroskedasticity hidden in the monitoring data, thereby further improving the accuracy of damage identification. Therefore, the proposed method may provide an innovative measure to assess the bridge structural condition in practice.

Important Links:



Go Back