东南大学学报（自然科学版）2025，Vol.55Issue(1)：59-65,7.DOI:10.3969/j.issn.1001-0505.2025.01.007
结构抗震可靠度分析的特征函数卷积方法
Convolution of characteristic functions for structural seismic reliability analysis
摘要
Abstract
To enhance the accuracy of extreme value distribution tail estimation,a novel method for structural seismic reliability analysis is proposed,employing characteristic function convolution techniques to achieve improved precision.The approach starts by expressing the extreme value distribution as a weighted Gaussian mixture distribution and derives an analytical expression for its characteristic function.By applying convolu-tion techniques to the numerical characteristic function,the method efficiently estimates all parameters in the analytical expression,including weight coefficients,means,and standard deviations.These parameters are then used to reconstruct the extreme value distribution and its cumulative distribution function,allowing for the calculation of structural seismic reliability.A numerical example validates the effectiveness of the method in reconstructing extreme value distributions.Comparisons with the Mellin transform and generalized distribu-tion reconstruction methods demonstrate the superior accuracy of the proposed method in extreme value distri-bution tail estimation.关键词
结构抗震可靠度/卷积/特征函数/极值分布/自适应Key words
structural seismic reliability/convolution/characteristic function/extreme value distribution/adaptive分类
建筑与水利引用本文复制引用
张付泰,徐军,兰昱璁..结构抗震可靠度分析的特征函数卷积方法[J].东南大学学报（自然科学版）,2025,55(1):59-65,7.基金项目
国家自然科学基金资助项目(52278178). （52278178）
