Open Access Open Access  Restricted Access Subscription or Fee Access

Investigation on Mapping Relationship Between Deflection and Cable Tension by Artificial Neural Networks



Cables usually serve as critical load-bearing structural components in cable-stayed bridges. Vertical deflection is an important overall indicator that directly reflects the health status of bridge. The relationship between cable tension and vertical deflection also reflects the underlying mechanism of structures. In this context, a machine learning-based approach is proposed. Cable tension and vertical deflection are caused by prestress, temperature and vehicle loads. In this study, we focus on the cable force and vertical deflection caused by vehicle loads. Artificial neural network (ANN) is employed as the supervised learning method to learn the relationship between cable tension and vertical deflection. Field monitored data of Jintang Bridge are used to test the validity of the proposed method. The results show that ANN is capable of modelling vertical deflection from cable tension data in the case of arbitrary vehicle distribution. The proposed method which integrates the local and global variables by using two different kinds of monitoring data, avoids the complicated mechanical derivations, and is proved to be environment-insensitive.


Full Text:



  • There are currently no refbacks.