卡尔曼滤波器
控制理论(社会学)
平滑的
白噪声
风力工程
侧风
加速度计
风速
流离失所(心理学)
工程类
计算机科学
结构工程
气象学
物理
航空航天工程
人工智能
计算机视觉
心理治疗师
电信
操作系统
控制(管理)
心理学
作者
Chang Yin,Xiongjun Yang,Lijun Liu,Sen Yang,Ying Lei
摘要
Current wind load identification methods are mainly based on structural acceleration responses, which require the installation of many accelerometers on structures, to identify fluctuating wind loads that are assumed as independent white noise processes. With the development of machine vision, structural displacement responses under wind loads can be noncontact observed. In this paper, an identification method is proposed for full‐field wind loads including both the fluctuating and mean wind components on buildings using only structural displacement observations. Wind loads are treated as unknown forces on buildings without the assumptions of independent white noise processes of fluctuating wind loads in current identification methods. To reduce the number of independent unknown wind loads to be identified, the spatial correlations of wind loads are first analyzed. Then, as displacement observation equation does not contain the unknown forces, the smoothing Kalman filter under unknown input without direct feedback (Smoothing KF‐UI‐WDF) algorithm is used for wind load identification. To validate the effectiveness of the proposed method, both stationary and nonstationary wind loads on a 20‐floor shearing building are identified, and the identification results are validated in both time and frequency domains.
科研通智能强力驱动
Strongly Powered by AbleSci AI