脉动风速随机过程的正交展开

Orthogonal expansion of stochastic processes for wind velocity

根据脉动风速功率谱的两种不同统计方法,定义了脉动风速Davenport谱的等价功率谱。在此基础上,构造了虚拟脉动风位移随机过程。利用Karhunen-Loeve分解的基本原理,对虚拟脉动风位移随机过程进行正交展开,进而获得脉动风速随机过程的正交展开表达式。研究的主要目的在于:希望用少量的独立随机变量来描述脉动风速随机过程的主要概率特性,为工程结构的抗风动力响应分析及动力可靠性研究奠定了基础。实例分析表明,本文的研究思路是可行的。

According to different statistical approaches, a stochastic process of pseudo-wind displacement was defined which is equivalent to Davenport^,s power spectrum. Utilizing the basic principle of the Karhunen-Loeve decomposition for stochastic processes, the orthogonal expansion of the pseudo-wind displacement was carried out and a series expression of the wind velocity fluctuations can be obtained accordingly. From this expression, the main probabilistic characteristics of the wind velocity fluctuations may be captured with only a few random variables, and therefore a foundation for the dynamic response analysis and reliability assessment of wind-excited buildings may be established. The accuracy and effectiveness of this orthogonal expansion approach is demonstrated by the close match of the obtained sampling power spectrum to Davenport^,s spectrum.