Dynamic displacement estimation of railway beam bridge based on integration of strain and acceleration

Abstract

Accurate dynamic displacement monitoring is critical for assessing structural safety of railway beam bridges, where conventional estimation methods face dual challenges in capturing displacement containing both quasi-static components affected by baseline drift and high-frequency components contaminated by noise. This study proposes a heterogeneous data fusion framework that synergizes strain and acceleration measurements to achieve accurate displacement estimation without requiring modal analysis. First, the dynamic displacement at the target point is derived from the measured longitudinal strain through the moment–curvature relationship based on Euler–Bernoulli beam theory, allowing for an accurate strain-derived displacement. Subsequently, the Kalman filtering algorithm is employed to fuse the strain-derived displacement and acceleration-integrated displacement, effectively addressing baseline drift and noise amplification. Finite element simulations of a 25 m simply supported beam bridge demonstrated that this method accurately estimates the displacements with both high-frequency and quasi-static components under various loads. Field tests on a simply supported bridge of Wuhan Metro Line 1 showed a tiny measurement uncertainty under 0.3 mm during train passages, validating the applicability of this method for practical application.