大跨屋盖围护结构风荷载极值概率模型的优化与对比研究

项目名称： 大跨屋盖围护结构风荷载极值概率模型的优化与对比研究

项目编号： No.51508184

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 建筑环境与结构工程学科

项目作者： 李寿科

作者单位： 湖南科技大学

项目金额： 23万元

中文摘要： 大跨屋盖主体结构很少出现风毁现象，其围护结构损毁是风致破坏的主要形式。大跨屋盖围护结构抗风设计常以极值风荷载为依据，合理的极值概率模型是预测围护结构极值风荷载的先决条件。本项目拟对广义极值分布模型、超阈值模型和转换过程极值模型进行优化与适用性研究。首先，采用刚性模型测压风洞试验方法，研究大跨屋盖风压系数的分布规律。其次，基于贝叶斯理论，考虑参数不确定性的影响，对广义极值分布模型的参数估计方法进行优化研究；基于变点理论和多变量相关过程非高斯仿真方法，研究阈值的量化自动选择，实现超阈值模型在小容量样本极值估计中的应用；采用Askey-scheme正交几何级数，研究转换过程极值模型在全峰度和偏度范围内的应用理论。最后，对上述三种模型进行系统的比较研究，针对不同的样本特性，研究其对应的最优极值概率模型，建立大跨屋盖围护结构设计风荷载极值概率模型的定量选择机制，为大跨屋盖围护结构设计提供理论参考。

中文关键词： 大跨度屋盖；围护结构；极值风荷载；概率模型；优化

英文摘要： Main Wind-Force Resisting System of long span roof is rarely destroyed by strong wind, but its claddings are usually damaged in wind disaster. The extreme wind loads are used as the design basis to resist the wind loads for claddings. So the reasonable probability distribution is most important for deciding the extreme wind loads on claddings. In this project, generalized extreme value model, peak over threshold model, and translation model are specially studied and optimized. First, the variation rules of wind pressure coefficients on long span roof are studied by wind tunnel tests. Secondly, based on the Bayes theory, the parameter estimation method is optimized for taking into account the uncertainty of parameters; based on the change point theory, the threshold is automatically chosen, and making the peak over threshold model to adapt minority samples by the multi-variable non-Gaussian simulation; using Askey-scheme hyper-geometric orthogonal polynomials to study the translation model for satisfying all the kurtosis and skewness of wind loads. Finally, a special contrastive study is carried out among all the probability distributions of extreme wind loads, and the selection mechanism is proposed to decide the best extreme probability distribution for designing claddings. This project is used as a reference resource for claddings design.

英文关键词： Long span roof;Cladding;Extreme wind load;Probability distribution;Optimization