中文摘要：

考虑系统参数的随机性，将基于广义卡尔曼滤波的子结构法与贝叶斯更新方法相结合，提出了桥梁结构基于贝叶斯更新物理参数的剩余强度估计两步法：第一步，将子结构法与广义卡尔曼滤波算法相结合，成功识别出子结构及其相邻单元的物理参数；第二步，视识别出的结构物理参数为更新信息，对以蒙特卡罗仿真实验结果作为先验分布的参数进行贝叶斯更新并分别基于蒙特卡罗仿真参数和贝叶斯更新物理参数对结构进行了剩余强度估计。数值算例表明：基于贝叶斯更新物理参数估计得到的结构剩余强度明显低于基于蒙特卡罗仿真参数估计得到的结构剩余强度。该方法为测量响应信息不完备条件以及小样本抽样情况下桥梁结构剩余强度估计提供了一个较好的解决思路。

英文摘要：

Taking the randomness of system parameters into account, this paper presents a two-step remaining load capacity estimation for bridge structures based on the Bayesian updated physical parameters. In the first step, the sub-structure method and the Extended Kalman Filter algorithm are combined to identify the physical parameters of the substructure and the adjacent elements. In the second step, the Bayesian updating process is taken and the identified physical parameters are treated as the new conditional distribution information while the Monte-Carlo simulation parameters are treated as the prior distribution information. Then, the remaining load capacity estimation is procured based on the Monte-Carlo simulation parameters and the Bayesian updated physical parameters respectively. The numerical examples show that the remaining load capacity estimation based on the Bayesian updated physical parameters is much lower than that estimated using the Monte-Carlo simulation. The method provides a good solution for remaining load capacity estimation of bridges with incomplete response information and small samples.