基于平稳小波包分解和希尔伯特变换的故障特征自适应提取  

Adaptive Fault Feature Extraction Based on Stationary Wavelet Packet Decomposition and Hilbert Transform

作　　者：刘毅华[1] 王媛媛[1] 宋执环[1,2] 

机构地区：[1]浙江大学宁波理工学院,宁波315100 [2]浙江大学控制科学与工程学系,杭州310027

出　　处：《电工技术学报》2009年第2期145-149,157,共6页Transactions of China Electrotechnical Society

摘　　要：提出一种自适应地提取信号特征分量的故障检测方法。采用逐层推进的平稳小波包分解算法,运用希尔伯特变换,在对信号进行小波包分解的同时,对分解结果进行瞬时频率和瞬时幅值分析,根据设定的分量提取和信号分解规则,实现信号分解路径的自主搜索,自适应地构建信号的小波包分解树,对信号进行多分辨率的频谱分析,达到信号消噪和特征分量提取的目的。仿真研究表明该方法的分量提取规则简单、目标明确,信号分析结果简洁,具有运算时间少、数据存储量小的特点和良好的抗噪性能,所提取的故障特征分量的时-频-幅值信息清晰、易于检测。A fault detection algorithm is proposed to adaptively extract the characteristic component of fault signals in this paper. The algorithm used one-level stationary wavelet packet transform to decompose the signal into low-and high-frequency subbands, whose instantaneous frequency and instantaneous amplitude are calculated by Hilbert transform at the same time. Based on the preset rules of component extraction and signal decomposition, the algorithm adaptively selects the path of stationary wavelet packet decomposition, building the wavelet packet decomposition tree of the signal with a multiresolution spectral analysis. Thus the algorithm denoiseds the signal and extracts the characteristic components for fault detection. The simulation show that the algorithm provides sufficient frequency-amplitude fault information with the less computational workloads and data storage spaces. The algorithm also has a good anti-noise performance.

关 键 词：平稳小波包变换 希尔伯特变换 自适应信号分析 多分辨率频谱分析 故障检测 

分 类 号：TM711[电气工程—电力系统及自动化]