建筑结构学报2026，Vol.47Issue(8)：246-258,13.DOI:10.14006/j.jzjgxb.2025.0516
建筑结构活荷载的大数据建模
Big data-empowered live load modeling for building structures:a case study of residential buildings in Shanghai
摘要
Abstract
Live load determination forms the basis for the design of building structures.Traditional modeling method of the live load is characterized by oversimplifies models,insufficient data support,and approximate combination methods.In view of this,this paper proposes a big data-empowered modeling method for the live load in building structures.A compound Poisson process is employed as the basic live load model.Internet big data and artificial intelligence support automated online investigation of live loads.The load coincidence principle and probability density evolution method enable accurate combination of multiple live loads,thereby achieving the live load modeling for building structures.Taking the residential buildings in Shanghai as an example,the proposed method is applied to a live load investigation covering 840 000 square meters,in which the scale of measured data far surpasses that of previous investigations.Based on this,the latest statistical results of the sustained load and extraordinary load are provided.A comparison between the latest and previous statistical results indicates that the data obtained in the 1970s fail to reflect the current situation of live loads in residential buildings.The relative difference of the statistical result describing the sustained load exceeds 40％.The probability distribution of the maximum combined live load within 50 years is calculated based on the previous modeling method and the proposed big data-empowered modeling method,respectively.The results indicate that the previous modeling method is not conservative,with the mean and standard deviation of the maximum combined load being about 30％and 50％lower,respectively.The big data-empowered modeling method is recommended in the live load modeling for building structures.关键词
活荷载/大数据建模/荷载模型/荷载调查技术/荷载组合Key words
live load/big data modeling/load model/load investigation technique/load combination分类
建筑与水利引用本文复制引用
徐迟,陈隽,李杰..建筑结构活荷载的大数据建模[J].建筑结构学报,2026,47(8):246-258,13.基金项目
国家自然科学基金重点项目(51538010). （51538010）
