A reliable methodology to estimate cable tension force in cable-stayed bridges using Unmanned Aerial Vehicle (UAV)

Abstract

An alternative methodology to estimate the tension force of cables using computer vision and Unmanned Aerial Vehicles (UAV) was proposed in this investigation. It was developed according to a simplified model to determine the fundamental frequency of the cables when are affected by daily vehicle loads. The methodology offers the advantage of accurately determining the tension force using just 2 min of video recording per cable, even under real environmental conditions. In addition to this, the calibration process of the camera can be Skipped. The proposed methodology was tested in the laboratory with a vibrating table and on the Rio Papaloapan Bridge, which is one of the main cable-stayed bridges in Mexico. The results in a controlled environment showed a Root Mean Square Error (RMSE) of 0.0024 Hz. Meanwhile, the assessment of the fundamental frequency obtained by the UAV and accelerometer in the structure showed a maximum difference of 0.03061 Hz, and a minimum of 0.00105 Hz. These discrepancies occasioned errors in tension force of 40.7 kN (1.96 %), and 3.3 kN (0.123 %), respectively.