Dynamic Prediction of the Remaining Useful Life of Bridge Bearings Jointly Driven by Monitoring and Inspection Data

The bearing slide plate is considered a loss component in bridges and will wear under the long-term action of external loads. The wear of bearings can lead to changes in the boundary conditions of a bridge, alter the internal forces of the structure, cause damage to structural components, and pose a threat to the structural safety. To ensure the bridge’s safe operation, it is necessary to accurately estimate the remaining useful life (RUL) of the bearing slide plate. Therefore, this paper proposed a dynamic prediction method for the RUL of the bearing slide plate, considering the change of girder end motion characteristics by integrating bearing monitoring and inspection data. First, based on Achard’s wear model, the quantitative relationship between bearing displacement and slide plate wear thickness was established, considering the effects of bearing pressure and sliding speed. The inspection data of the slide plate wear thickness were used to eliminate the effects of variation in lubrication conditions and displacement measurement distortion due to the low sampling frequency of the sensor. Then, considering the uncertainty and time variability of the slide plate wear process during the bridge’s service life, the RUL prediction model for the bearing slide plate was established based on the inverse Gaussian stochastic process. Based on the Bayesian theory and the expectation maximization algorithm, the prediction model parameters were dynamically updated by using the monitoring data of the bearing displacement, and the bearing RUL was dynamically predicted based on the monitoring data. Finally, a long-span bridge case study demonstrates that the proposed method can accurately predict the RUL of the bearing slide plate, providing guidance for developing a repair and replacement strategy.