A data assimilation method about Bayesian Fourier dynamic linear prediction of periodic extreme stresses for steel bridges

Abstract Dynamic prediction of periodic extreme stresses for steel bridges facilitates bridge preventive maintenance decision-making. This paper proposes a data assimilation method for predicting periodic extreme stresses of steel bridges. To this aim, Fourier dynamic linear model (FDLM) is firstly built with Fourier function, Taylor series expansion technology and dynamic linear models; secondly, dynamic probability recursive processes of FDLM are provided with Bayesian method, further, the periodic extreme stresses can be dynamically predicted based on the monitoring data; finally, the monitoring periodic extreme stress data of a steel bridge is provided to illustrate the effectiveness of the proposed method, the prediction results show that the proposed method is able to not only predict dynamic extreme stresses with periodicity, but also obtain superior prediction accuracy compared with the existing methods.