A Kalman filter algorithm for identifying track irregularities of railway bridges using vehicle dynamic responses. (April 2020)

• Record Type: Journal Article
• Authors: Xiao, Xiang; Sun, Zhe; Shen, Wenai
• Is Part Of: Mechanical systems and signal processing. Volume 138(2020)
• Publication Date: 2020-04
• Subjects: Track irregularity identification -- Railway track monitoring -- Kalman filter -- Time-dependent system -- Vehicle-bridge interaction -- Unknown input; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2019.106582
• Languages: English
• ISSNs: 0888-3270

Abstract

Highlights: An algorithm is proposed to identify track irregularities using vehicle responses. TThe VB interaction is taken into account in the proposed algorithm. A railway bridge is employed as an illustrative example to validate its accuracy. Abstract: Track irregularities affect the running safety of railway vehicles and ride comfort, hence track irregularity identification using the dynamic responses of in-service vehicles is of great interest. Because the high-speed rail lines mainly consist of bridges in China, vehicle-bridge (VB) interactions which significantly influence the vehicle dynamic responses should be taken into account in the track irregularity identification. This paper proposes a Kalman filter algorithm to identify the track irregularities of railway bridges using vehicle dynamic responses considering the VB interactions in real-time. A state space model is established to represent a time-dependent VB system subjected to unknown track irregularity excitations. A Kalman filter algorithm is proposed to estimate optimally the state vector of the VB system and to identify the track irregularities subsequently. Two numerical examples including a real railway bridge constructed in China are presented to validate the accuracy of the proposed algorithm. A parametric study is also conducted to demonstrate the effects of measurement noise, vehicle running state, parameter uncertainty and model uncertainty on the identification of track irregularities. Comparison results demonstrate that the proposed track irregularity identification algorithm outperforms the conventional approaches mainly because of considering the VB interaction. The proposed algorithm enables efficient monitoring the track irregularities of railway bridges using the acceleration responses of in-service vehicles.