振动信号的多尺度分解、三维重构与识别  

Multi-Scale Decomposition, 3-Dimension Reconstructional and Identification of Vibration Signals

作　　者：王新晴[1] 王耀华[1] 唐建宁[1] 陆明[1] 徐燕申[2] 赵相松[2] 

机构地区：[1]解放军理工大学工程兵工程学院,南京210007 [2]天津大学机械工程学院,天津300072

出　　处：《振动工程学报》2001年第4期473-476,共4页Journal of Vibration Engineering

摘　　要：小波多分辨分析能够实现振动信号的任意精度与尺度的时 -频域分解 ,而改进的 Choi- Willons分布实现信号的三维重构 ,并有效地揭示典型信号的分布特征。实例证明 :两种方法的结合对于在强噪声等复杂背景下 ,提取与识别非平稳。Wavelet multi-scale analysis can decompose vibration signals at any scale and accuracy. And modified Choi-Willons distribution can reconstruct signals in 3-dimension coordinates, in which every typical signal component has different distribution features. Then the algorithm and the examples of the decomposition, denoising, reconstruction and identification of vibration signals are analyzed in details. It is confirmed that the combination of the two methods plays important and effective role in solving lots variety of problems, such as revealment and identification of unstable and transient components in the background of mighty noise and disturbances.

关 键 词：多尺度分解 三维重构 小波多分辨分析 时-频分布 振动信号 信号识别 

分 类 号：TB53[理学—物理] TN911.7[理学—声学]