台风非平稳特性分析及全桥风场模拟  

Analysis and simulation of non-stationary typhoon field on long-span bridge

作　　者：郝键铭 冯宇[1] 苏权科 韩万水[1] HAO Jian-ming;FENG Yu;SU Quan-ke;HAN Wan-shui(School of Highway,Chang’an University,Xi’an 710064,Shaanxi,China;Hongkong Zhuhai Macao Bridge Authority,Zhuhai 519015,Guangdong,China)

机构地区：[1]长安大学公路学院,陕西西安710064 [2]港珠澳大桥管理局,广东珠海519015

出　　处：《长安大学学报（自然科学版）》2022年第6期66-76,共11页Journal of Chang’an University(Natural Science Edition)

基　　金：国家重点研发计划项目(2019YFB1600702);国家自然科学基金项目(51878058);中央高校基本科研业务费专项资金项目(300102210108)。

摘　　要：为了实现基于数据驱动的跨海桥梁非平稳台风风场模拟,采用基于离散小波变换(DWT)的非平稳风速模型提取台风时变平均风速,并分析比较了采用平稳模型和非平稳模型获得的平均风及脉动风特性,强调了台风显著的非平稳特性;采用小波包变换(WPT)和Hilbert变换提取非平稳信号的瞬时信息,通过逆变换实现单点非平稳风速信号模拟;基于时频相干函数建立空间多尺度相关系数矩阵,通过相关系数矩阵Cholesky分解建立频率载体函数的空间相关性,实现多点非平稳风速时程模拟,并应用于某跨海斜拉桥全桥非平稳风场模拟。研究结果表明:采用平稳和非平稳模型计算的脉动风特征参数具有一定差异性,这种差异性随着基本时距的减小逐渐减小;Hilbert谱可以直观展现台风风速能量的时频分布特征,刻画非平稳脉动风瞬时特征;某跨海斜拉桥全桥非平稳风场模拟结果与实测值的Hilbert谱、Arias强度、Rényi熵、分形维数吻合较好,与目标空间相关系数吻合较好,验证了模拟结果的可靠性。To realize the non-stationary typhoon wind field simulation of sea-crossing bridge based on data, non-stationary wind speed model based on discrete wavelet transform(DWT) was used to extract time-varying mean wind, the characteristic of mean wind and fluctuating wind obtained from stationary model and non-stationary model were analyzed, with the non-stationary characteristics of typhoon emphasized. The wavelet packet transform(WPT) and the Hilbert transform was used to obtain instantaneous information of non-stationary signal, univariate non-stationary wind speed was simulated by inverse transformation. The spatial multiscale correlation coefficient matrix was established based on coherence-correlation function, correlated frequency-modulated subcomponents was generated from uncorrelated frequency-modulated subcomponents using Cholesky decomposition. Hence, multivariate non-stationary wind speed could be simulated, and it was applied to the non-stationary wind field simulation of a sea-crossing cable-stayed bridge. The results show that the significant difference between the turbulence characteristics of fluctuating wind obtained from stationary model and non-stationary model can be observed, which decreases with the decrease of averaging interval. Hilbert spectrum can intuitively show the time-frequency distribution characteristics of typhoon wind speed energy and describe the transient characteristics of non-stationary fluctuating wind. The non-stationary wind field simulation of a sea-crossing cable-stayed bridge are in good agreement with measurements in terms of Hilbert spectrum, Arias intensity, Rényi entropy, fractal dimension and the correlation coefficient, which means the simulation is valid. 13 figs, 24 refs.

关 键 词：桥梁工程 台风模拟 HILBERT变换 小波包变换 非平稳 跨海桥梁 

分 类 号：U442[建筑科学—桥梁与隧道工程]