基于K-L展开的高耸结构上分布随机风载荷识别

Identification of distributed stochastic wind load on high-rise structures based on the K-L expansion

针对结构上分布式风载荷难以直接测量的问题,给出了一种利用位移响应,基于模态空间转换和Karhunen-Loéve(K-L)展开的分布风荷载间接识别方法。采用Davenport谱模拟风荷载的随机特征,计算高耸结构在有限测点处的动位移响应作为风荷载识别的"实测"输入。将风荷载的分布函数利用结构模态振型函数展开,从而将包含无限个时间历程的分布动荷载识别问题转换为多参数的动荷载识别问题。通过"实测"动响应样本集的K-L展开识别风荷载的分布特征以及统计特性。以一桅杆结构为对象开展数值仿真,验证了所提出识别方法的有效性。

To deal with the difficulty of directly measure the distributed wind load on structures, an indirect distributed wind load identification method based on modal space conversion and Karhunen-Loéve (K-L) expansion using displacement response was proposed. The Davenport spectrum was adopted to simulate the stochastic characteristics of the wind load and the displacement responses at limited locations were calculated as the “measured” input for wind load identification. The distribution function of the wind load was expanded by the structural mode shape functions, which transforms a distributed dynamic force identification problem with infinite time histories to a dynamic load identification problem with multiple parameters. Via the K-L expansion of the “measured” response sample sets, the distribution and statistics can be identified. Numerical simulations were conducted on a guyed mast structure which proofs the effectiveness of the proposed identification method.