Stiffness Evaluation of Multigirder Bridges Using Vision‐Based Displacement Measurement and an Iterative Algorithm

Prefabricated multigirder bridges offer significant advantages and have been extensively adopted for medium‐ and small‐span highway bridges in China. Owing to the complex spatial load distribution among individual girders, damage detection and quantification for such bridges remain a challenging problem. Damage alters the internal force distribution, thus making the conventional transverse load distribution model based on uniform stiffness assumptions invalid. This paper proposes a novel damage identification framework that integrates vision‐based structural monitoring with an iterative matrix displacement algorithm for stiffness estimation. First, traffic video streams are analyzed using computer vision algorithms to detect, classify, and track vehicles, while rotational responses are simultaneously measured at multiple locations along the bridge. The spatiotemporal distributions of vehicle weight are then synthesized by fusing trajectory data with toll station records. Second, a two‐level detection strategy is proposed to locate and quantify stiffness reduction via an iterative algorithm. A new damage indicator defined as the ratio of the bending moment envelope area (BEA) to the curvature envelope area (CEA) is formulated to localize structure damage (Level 1); a matrix displacement‐based iterative algorithm is then developed to quantify stiffness degradation precisely (Level 2), which accounts for internal force redistribution effects by progressively updating the regional stiffness parameters. Finally, the proposed framework is validated on a scaled multigirder bridge model under controlled damage scenarios, demonstrating its effectiveness in both damage localization and severity estimation.