Digital twin-based temperature effect separation from responses of an actual cable-stayed bridge

Field structural response measurements often exhibit non-negligible variations due to environmental temperature fluctuations. Due to this, a digital twin (DT) based framework has been proposed for separating temperature effects from the measured responses of an actual cable-stayed bridge by providing a system-level analysis perspective. A cable force influence matrix serves as the information interaction medium by deriving adjustment vectors that transform the monitored cable forces into the modelling information required for the DT modelling, which ensures a precise alignment between the simulated and measured responses. The DT model can then treat the temperature variations as input environmental loads, realising the simulation and quantification of the temperature-induced effects on the structural responses of the physical bridge. The feasibility of the proposed framework has been verified using the monitored cable forces and temperatures of the cable-stayed bridge. It was found that the DT model could timely evolve in conjunction with the monitored temperatures, accurately mirroring the measured cable forces. The analysis results also indicated that the temperature variations significantly affected the cable forces, with a greater sensitivity observed in the longer cables, which was demonstrated by a strong correlation coefficient of 0.871 for the longest cable.