Summary
From 2010 to 2016, wind speed time history and synchronous acceleration responses along Xihoumen Bridge (XB) girder were recorded by a health‐monitoring system. Based on this dataset, wind speed, wind direction, angle of attack (AoA), turbulence intensity, and turbulence length scale are studied based on 10‐min time intervals. The environmental wind characteristics and synchronous structural effects are decoupled by the control variable method. The probabilistic dependence between wind parameters and structural responses is explored by explicit polynomial regression equations. With wind environment parameters as input, the acceleration responses of the bridge could be rapidly predicted with a certain confidence interval using decoupling and regression modeling (DRM). The proposed DRM can reproduce the buffeting acceleration responses when the bridge is subject to future potential strong typhoon attacks. The key factors affecting the buffeting responses of XB are analyzed and discussed. Generally, the buffeting acceleration responses of XB are significantly affected by wind speed and turbulence intensity rather than by turbulence length scale, which agrees well with previous observations.