STUDY ON THE EXTREME VALUE ESTIMATING METHOD OF NON-GAUSSIAN WIND PRESSURE

全涌,顾明,陈斌,田村幸雄

2010-01-01

Abstract:Based on the classical extreme-value theory, an estimating method for expected extreme-value of wind pressure with non-Gaussian probability distribution is proposed with a sample whose length is just one standard observation interval. At first, the wind tunnel test and test data process of this study are introduced in detail. A method to calculate expected extreme values of time series with a long time interval is then proposed with its sub-sections based on the classical Gumbel theory for extreme values and the independence of observed extreme values. At last, the extreme values of the wind pressure coefficients of the present wind tunnel test are calculated with the proposed method and methods used widely at present, such as peak factor method, improved peak factor method and Sadek-Simiu method. The results indicate that the probabilistic parameters of extreme values of sub-sections of a non-Gaussian wind pressure time series can be used to estimate accurately expected extreme value of their parent section with the proposed method. The length of the sub-section can be determined with auto-correlation analysis on the parent section. Comparison shows that the proposed method can estimate extreme values of non-Gaussian wind pressure more accurately than other methods used widely at present.