空气动力学
控制理论(社会学)
时域
解耦(概率)
情态动词
气动弹性
频域
空气动力
计算机科学
风洞
滤波器(信号处理)
声学
工程类
物理
控制工程
计算机视觉
航空航天工程
人工智能
化学
高分子化学
控制(管理)
作者
Xiaoqi Hu,Zhong Xie,Lele Zhang
出处
期刊:Journal of Engineering Mechanics-asce
[American Society of Civil Engineers]
日期:2021-10-01
卷期号:147 (10)
被引量:1
标识
DOI:10.1061/(asce)em.1943-7889.0001990
摘要
The high-frequency force balance (HFFB) technique is commonly used in wind tunnels to investigate the wind loads and wind-induced responses of high-rise buildings. This paper presents a time-domain correction method for the signal distortion in HFFB tests, including the coupling effect among the balance components and the dynamic amplification of the balance-model system (BMS) on aerodynamic loads. The proposed method uses an adaptive blind source separation (BSS) method to decouple the measured aerodynamic loads in real-time and a fitting method considering the actual aerodynamic characteristics to conduct modal parameter identification of the decoupled modal signals. Subsequently, a digital compensation filter is constructed on the basis of the identified parameters, and the decoupled modal signals are corrected in the time domain by the filter. Finally, the corrected modal signals are transformed into physical signals by using the mode shape, and the time history of the corrected aerodynamic loads is obtained. The proposed method can update the separating matrix (i.e., the inverse of the mode shape) of the coupled signals online and accomplish the real-time decoupling. Besides, the time history of the corrected aerodynamic loads can be directly obtained using the filter-based correction method, which is convenient for further time history analysis. The proposed method is applied to both a numerical simulation example and an HFFB test and compared with other existing methods. The proposed method reduces the error maximum and the standard deviation by 81.9% and 88.9%, respectively, in the numerical simulation. The results show the effectiveness and superiority of the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI