摘要
针对从受噪声污染的脉冲响应信号和快速正弦扫频响应信号中识别振动系统的单模态和多模态的频率问题,将小波变换与奇异值分解(SVD)相结合,利用基于小波变换的能量分布函数,进行奇异值分解后,根据最大奇异值,求出主分量,然后通过主分量对应的尺度图或频率计算公式,识别出模态的频率。仿真结果证明了该方法的有效性和可行性。
In accordance tO identifying the single mode and closely modes frequency of the vibration systems by impulse response and the quick sinusoidal sweep response with serious noise, wavelet transform was combined with SVD filtering. Firstly, energy distribution function based-on wavelet transform was used in singular value decomposition, and then primary component was extracted on the basis of maximum weight, and finally mode frequency was identified according to scalograms of the primary component or frequency formula. Simulation results show that the method is feasible and effective.
出处
《系统仿真学报》
EI
CAS
CSCD
北大核心
2006年第8期2227-2229,共3页
Journal of System Simulation
关键词
小波变换
尺度图
奇异值分解(SVD)
振动系统
模态频率识别
wavelet transform
scalograms
SVD（singular value decomposition）
vibration systems： identification of mode frequency
