识别复杂屋盖风压场的改进本征正交分解方法

Identification of wind pressure fields over a complicated roof by an improved proper orthogonal decomposition technique

为了提高本征正交分解(POD)方法识别复杂体型屋盖结构且结构表面测点布置不均匀的风压场的精度,提出了改进加权POD方法,采用双加权的手段,以结构上测点风压自根方差和测点面积做为权重.为了避免淡化某个权重的影响,首先将两个权重做了规格化处理.以深圳市民中心的屋盖结构为工程背景,根据风洞试验同步采集的风压数据,来识别作用在结构表面的风压.利用该改进方法,重建了深圳市民中心屋盖结构的随机风压场,对重建的风压时程、均值以及根方差值与实测风压时程、均值以及根方差做了比较,利用本文提出的方法识别典型测点的风压时程均值的最大误差为11.37%、根方差的最大误差为29.3%;而加权POD法识别典型测点的风压时程均值的最大误差超过100%出现了识别错误、根方差的最大误差达到了55.6%.

An improved weighted orthogonal decomposition technique was proposed to improve the estimation of air flow pressure fields over a complicated roof having nonuniformly distributed pressure taps. The area and root variation of the wind pressure taps were weighted and normalized. The roof of the Shenzhen Citizens Center was taken as an example and the wind pressure over the structure was estimated from wind tunnel tests. Random wind pressures on the roof were estimated using the method proposed in the paper. The estimated wind pressure over time, the average pressure, and the variance of the pressure were compared. The largest errors in the average and the variance were 11. 37% and 29.3% respectively. The weighted orthogonal decomposition method gave a maximum error that exceeded 100%, which was the result of a wrong identification. The largest error in standard deviation over time was 55.6%.