GPS-structural health monitoring of a long span bridge using neural network adaptive filter

The movement of bridge deck bearings plays a significant role in the safety of bridges. Real time kinematic global positioning system (GPS) continuous health monitoring using relative deformations was carried out on a long span Zhujiang Huangpu Bridge. The neural network aided adaptive filter is used to predict and adjust the GPS monitoring data. The statistical moments in time and frequency domains were used to analyse the movement of the bridge deck. The results indicate that (1) the proposed neural network with the adaptive filter model can be used to de-noise the GPS health monitoring signals, (2) the GPS is highly sensitive for bridge deck movements, (3) the statistical moments can be used to detect the movements and errors of the GPS observations, and (4) the bridge is very safe under different loads.