考虑不确定性斜拉桥参数的重要性与易损性分析

许多结构参数对桥梁数值模型的计算结果产生影响，结构的恒载、阻尼以及相关材料特性等参数涉及的不确定性被称为建模不确定性。研究表明建模不确定性的存在使得结构易损性曲线的偏差程度达到70%，因此在地震易损性分析中考虑不确定性很有必要。文章采用一种基于随机森林（random forest，RF）模型的概率地震需求模型（probabilistic seismic demand model，PSDM）易损性计算方法，分析桥梁构件的材料特性、荷载条件和构件几何特性等不确定因素对桥梁易损性的影响，并从中筛选出重要的不确定性参数。借助随机森林模型的学习与预测功能对结构易损性曲线进行预测。

Many structural parameters, such as the dead load, damping and related material properties of the structure, will affect the calculation results of the bridge numerical model. The uncertainties involved in these parameters are called modeling uncertainties. The research shows that the deviation of the structural vulnerability curve reaches 70% due to the modeling uncertainties, so it is necessary to consider the uncertainty in the seismic vulnerability analysis. This paper uses a probabilistic seismic demand model (PSDM) vulnerability calculation method based on random forest (RF) model to analyze the impact of uncertain factors such as the material characteristics, load conditions and geometric characteristics of bridge components on the bridge vulnerability, and screens out important uncertain parameters. The structural vulnerability curve is predicted by the learning and prediction function of RF model.