围护结构非高斯风压极值估计的改进独立风暴法

发布时间：2018-09-14 12:47

【摘要】：为获得具有确定保证率的围护结构设计风荷载,提出了针对较小容量样本的极值概率分析方法——改进独立风暴法,基本思路是采用自相关分析技术提取单段风压时程的多个独立峰值,进而结合加权最小二乘法对独立峰值进行概率分析,获得样本极值的最优无偏估计。由于样本抽取时保证了各峰值的独立性,且各独立峰值的形成机制具有较好的一致性,因此应用这一方法可对较小容量样本获得较为精确的极值概率模型。结合多次采样风洞试验,将改进独立风暴法与峰值因子法、改进峰值因子法和Sadek-Simiu法进行了对比,结果表明,前者可对较小容量样本获得极值概率模型的解析表达,进而获得具有指定分位点的极值,从而提高了概率分析方法在实际围护结构抗风设计中的适用性。
[Abstract]:In order to obtain the design wind load of the enclosure structure with certain guarantee rate, an improved independent storm method is proposed for extreme probability analysis of small size samples. The basic idea is to use autocorrelation analysis technique to extract multiple independent peaks of single stage wind pressure time history, and then combine the weighted least square method to analyze the probability of independent peak value, and obtain the optimal unbiased estimation of sample extreme value. Because the independence of each peak value is guaranteed and the formation mechanism of each independent peak value is consistent in sample extraction, a more accurate extremum probability model can be obtained by using this method for smaller samples. Combined with the multiple sampling wind tunnel test, the improved independent storm method is compared with the peak factor method, the improved peak factor method and the Sadek-Simiu method. The results show that the former can obtain the analytical expression of the extreme probability model for the smaller sample. Furthermore, the extreme value with a specified locus is obtained, which improves the applicability of the probabilistic analysis method in the wind resistant design of practical envelope structures.
【作者单位】：北京交通大学土木建筑工程学院结构风工程与城市风环境北京市重点实验室;哈尔滨工业大学土木工程学院;
【基金】：国家自然科学基金项目(51308039,91215302) 高等学校学科创新引智计划(B13002) 中国博士后科学基金资助项目(2013M530021) 中央高校基本科研业务费(2013JBM057)
【分类号】：TU312.1

【参考文献】