Cable Force Identification and Finite Element Model Optimization of Cable-Stayed Bridges Based on Backpropagation Neural Networks. Issue 5 (May 2023)

• Record Type: Journal Article
• Authors: Ouyang, Ping; Shen, Quanxi; Xie, Xiaoli; Zhu, Wanxu
• Is Part Of: Transportation research record. Volume 2677:Issue 5(2023)
• Page Start: 579
• Page End: 589
• Publication Date: 2023-05
• Subjects: bridges and structures -- construction -- infrastructure -- infrastructure management and system preservation -- monitoring -- structural health monitoring -- structure maintenance; Transportation -- Periodicals; Roads; Transport -- Périodiques; Routes -- Périodiques; Routes -- Conception et construction -- Périodiques; Roads; Transportation; 388.05
• DOI: 10.1177/03611981221133385
• Languages: English
• ISSNs: 0361-1981

Abstract

Cable force is an essential indicator for evaluating the health status of a bridge. To realize the real-time and accurate cable force monitoring of the whole bridge, models were constructed using backpropagation neural networks combined with a finite element model of a cable-stayed bridge. This strategy obtained the cable forces in the stay cables without sensors, the elastic moduli of the stay cables, and the elastic modulus of the bridge girder concrete. The results showed that the average differences in the forces in the 75 stay cables without sensors obtained from our identification model and those measured in 21 stay cables with sensors presented a maximum discrepancy of 0.17%. Then, the structural parameters from measured data were used to update the finite element model. All the results calculated via the cable force formula presented an error of about ±1% compared to the measured results. This research demonstrated that the models for identifying cable forces and bridge parameters provide a valuable and novel approach to force identification in stay cables without sensors.