The time-varying mean (TVM) component plays a vital role in the characterization of non-stationary winds, whereas it is difficult to extract the TVM accurately or to validate it quantitively. To deal with this problem, this paper first develops two additional conditions for the TVM extraction from the perspective of structural wind-induced vibration response, then presents an approach, based on the combination of Vondrak filter and genetic algorithm (Vondrak-G), to derive the optimal TVM from non-stationary wind speed records as well as its turbulence characteristics (i.e. gust factor, turbulence intensity, and turbulence integral length scale). Furthermore, the wind characteristics obtained by the Vondrak-G approach are compared with those by a conventional approach derived for stationary winds, demonstrating that the results by the Vondrak-G approach are evidently more accurate. This paper aims to provide an effective method for accurately extracting the TVM and then evaluating wind characteristics of the non-stationary wind.

中文翻译：

---

利用 Vondrak 滤波器和遗传算法提取非平稳风的时变平均分量：风工程视角

时变平均（TVM）分量在非平稳风的表征中起着至关重要的作用，而TVM难以准确提取或定量验证。针对这一问题，本文首先从结构风致振动响应的角度提出了两个额外的 TVM 提取条件，然后提出了一种基于 Vondrak 滤波器和遗传算法（Vondrak-G）相结合的方法，从非平稳风速记录及其湍流特征（即阵风因子、湍流强度和湍流积分长度尺度）中推导出最佳 TVM。此外，将通过 Vondrak-G 方法获得的风特性与通过传统方法导出的静止风的风特性进行比较，证明 Vondrak-G 方法的结果显然更准确。本文旨在提供一种有效的方法来准确提取TVM，进而评估非平稳风的风特性。