摘要
采用快速独立分量分析(FastICA)算法,快速、准确地识别结构模态参数.该算法以模态响应之间的独立性为依据构建出目标函数,并以此目标函数为基准采用ICA算法对结构输出信号进行分离,而得到结构振型向量.进而通过单模态识别技术-希尔伯特(Hilbert)变换,识别出结构的频率和阻尼比.最后通过白噪声激励下六层框架结构的模态参数识别,验证了快速独立分量分析算法识别结构模态参数的可行性和鲁棒性.
This paper presented a structural modal parameters identification by Fast Independent Component Analysis（ Fast ICA） rapidly and accurately. The method builds the objective function based on the independence between the modal responses,and separates the structural output signals by Fast ICA based on the objective function,and gains the structural modal shape vectors. Then structural frequencies and damping ratios are gained by Hilbert transform,a single mode analysis technical. Finally,the efficiency and robust of structural modal parameters identification by Fast ICA presented herein was verified by a six-story frame structure.
出处
《动力学与控制学报》
2016年第4期379-384,共6页
Journal of Dynamics and Control
基金
江苏省自然科学基金项目(BK20141180)
苏州科技学院科研基金项目(XKZ201304)~~
关键词
快速独立分量算法
模态参数识别
模态响应
目标函数
希尔伯特变换
fast independent component analysis
modal parameters identification
modal response
objective function
Hilbert transform
