|
Title |
Cable Damage Detection for Cable-Stayed Bridges Using Cable Force Variation Monitoring and System Identification Technique
|
|
Authors |
양기재(Ki-Jae Yang) ; 김정중(Jung-Joong Kim) |
|
DOI |
https://doi.org/10.11112/jksmi.2026.30.4.228 |
|
Keywords |
케이블 손상; 시스템 인식; 축소 SI; 사장교; 케이블 장력변동; 최적 센서 배치 Cable damage; System identification; Reduced SI; Cable-Stayed bridge; Cable force variation; Optimal sensor arrangemnet |
|
Abstract |
This study proposes a local cable damage detection method for cable-stayed bridges using cable force variation and System Identification (SI). Existing SI methods often suffer from overfitting and poor convergence due to field measurement noise and the low sensitivity of cable physical properties. To overcome this, a ‘Reduced SI’ technique was developed by classifying cables into four groups based on sag and bending stiffness parameters from 20 Korean cable-stayed bridges, selectively restricting identification variables to ensure fast and accurate tension estimation. Furthermore, a global SI algorithm was established by linking the identified tension variation to a global bridge finite element model (FEM). By combining L1-Norm regularization, Non-Negative Least Squares (NNLS), and Bayesian inference, the algorithm successfully pinpoints the location and cross-sectional reduction of damaged cables. Validation using field data from the HT Bridge with actual strand detachment confirmed that the proposed method identifies damage severity and location within a 10% error margin. Additionally, an optimal accelerometer placement strategy utilizing binary hypothesis testing is presented, establishing a practical framework for continuous structural health monitoring of in-service cable-stayed bridges.
|