Abstract: The service state of the steel deck system is closely related to load， temperature， structural response and other factors. In order to accurately analyze the long-term performance of the steel deck pavement， the study conducted long-term monitoring on an orthotropic steel deck system with cast + epoxy asphalt pavement and steel box girder， using a specific bridge as the engineering case. Based on a large number of real bridge monitoring data， the data analysis was carried out on service conditions and structural responses， such as temperature， vehicle speed， load， and characteristic values of structural response. The relations between different factors and the strain of steel deck-U-rib welds were studied and analyzed， and the multi-dimensional correlation rule of structural dynamic responses was established. The research showed that the strain amplitude of the weld seams in the cast+epoxy pavement structural steel deck-U-rib was linearly related to the axle load， and the strain magnitude increased as the axle load increased. For the same axle load， when the speed increased from 40 km/h to 70 km/h， the evaluated strain decreased by 28%。 When the temperature rose from 5 ℃ to 50 ℃， the average strain increased by 24%. Based on the real bridge monitoring data， a multi-factor regression model was established to quantify the relationship between weld strain and factors such as axle load， vehicle speed， and temperature. The model was then applied to predict strain variations in the weld seams.