大跨度屋盖风压场的条件模拟方法研究

Study on conditional simulation of wind pressure fields for large-span roofs

风荷载是大跨度屋盖的控制荷载。通常由于风洞实验测得的风压数据有限,那么利用有限的数据预测目标点的风压场就尤为重要。考虑到大跨度屋盖风压场的复杂性,本文提出了一种简单的非参数化条件模拟方法。该方法采用Karhunen-Love(简称KL)扩展将风压随机变量表示为协方差算子的本征函数。为降低变量维数,给出了KL扩展的转移分布的最邻近估计,然后通过扩展本征函数的Fourier基,并估计其扩展系数来实现对本征函数的估计,进而最终获得目标风压场。将该方法应用到大跨度结构的风压场估计中,得到了大跨度结构的风压场相关参数,计算结果与实验值符合良好,且结果表明本方法的计算效率高,误差小。这种方法有效地降低了所预测风压场变量的维数,在模拟大跨度屋盖风压场时达到了计算准确、误差小、效率高的目的。

Wind loading is the control load for large span roofs. Due to the limited wind pressure data obtained from wind tunnel tests, it is quite meaningful to predict the targeted wind pressure fields employing the limited measured data. Considering the complicated wind pressure field of large span roofs, a simplified non-parametric simulation method is proposed. Karhunen-Lo^ve （KL for short） expansion is adopted to represent wind pressure random variants as eigen-functions of covariance operator. The nearest neighbour estimator is presented for transi- tion distribution of KL expansion to reduce the variant dimensionality. Then targeted wind pressures fields are ob- tained through Fourier basis expansion of eigen-functions and estimating the expansion coefficients to perform eigen- functions estimation. The method is applied to estimate the wind pressure fields of a large span roof, and wind pres- sure field related parameters are obtained, which are coincident with those from tests, with small errors and high ef- ficiency. It is found that the proposed method effectively reduces the dimensionality of the predicted wind pressure fields, and reaches the goal of small errors, high accuracy and efficiency for simulating the wind pressure fields of large span roofs.