(1 长安大学公路学院， 西安 710064； 2 同济大学土木工程学院， 上海 200092；

3 中国建筑设计院有限公司， 北京 100044)

［摘要］建筑结构极值风速预测的研究在结构设计与安全使用领域具有重要意义。基于Bayes理论提出极值风速预测方法，采用Monte Carlo数值模拟方法产生服从极值-I型分布的伪风速母样，基于伪风速母样分别采用基于Bayes理论和最大似然估计理论的极值-I型风速预测方法进行风速预测，并与伪风速母样的理论值进行对比分析。研究结果表明：与最大似然估计法相比，采用基于Bayes理论建立的极值-I型风速预测模型进行风速预测的精度更高。

［关键词］Bayes理论； 极值-I型； 伪风速母样； 风速预测； 最大似然估计

中图分类号： TU973.2+13 文献标识码： A文章编号：  1002-848X(2016)S1-0986-03

Research on extreme wind speed forecasting method of building

 structure based on Bayesian theory

Lou Canhong1, Dong Fenghui2, Yao Yuan3, Lian Yu2

(1 Highway College of Chang′an University, Xi′an 710064, China; 2 School of Civil Engineering, Tongji University,

 Shanghai 200092, China; 3 China Architecture Design Group, Beijing 100044, China)

Abstract: The investigation of wind speed forecast of extreme value type distribution has a great significance for the building structural design and use. The wind speed forecast method of extreme value type distribution were constructed, and Monte Carlo simulation method is used to produce the pseudo wind speed samples. The maximum likelihood parameter estimation method and Bayes statistical theory were used to estimate the wind forecast value of the extreme value type I distribution, which has been done by comparing with the theoretical extreme value type I distribution of the pseudo wind speed. The result indicates that the wind speed forecast model of extreme value type I distribution based on Bayes theory is optimal than the maximum likelihood estimation. 

Keywords: Bayes theory; extreme value type I distribution; pseudo wind speed sample; wind speed forecast; maximum likelihood estimation