A novel Bayesian extreme value distribution model of vehicle loads incorporating de-correlated tail fitting: Theory and application to the Nanjing 3rd Yangtze River Bridge

•Monitored extreme values of vehicle loads disagree with the estimated ones.•We use a GPD tail distribution to improve the extreme value model’s accuracy.•We use Peak Over Threshold method to exclude correlation among vehicle loads.•We developed a Bayesian form of the above model and several applications.•The estimation results confirm the new model’s superiority.