基于小波包变换和关联维数的空化信号特征提取  

Feature Extraction of Turbine Cavitation Based on Wavelet Packet and Fractal Analysis

作　　者：李静[1] 周建中[1] 肖剑[1] 蒲桂林[1] 李超顺[1] 肖汉[1] 

机构地区：[1]华中科技大学水电与数字化工程学院,湖北武汉430074

出　　处：《水力发电》2013年第10期53-57,共5页Water Power

基　　金：国家自然科学基金项目(51079057;51039005;51109088)

摘　　要：在详细阐述了声发射监测技术及空化声发射信号的特征基础上,采用小波包变换方法对空化信号进行处理。引入小波包特征频段分析法,以特征频段的能量和关联维数作为特征量来表征空化信号特征。最后根据选择的特征量对不同工况下的空化进行分类。仿真实验结果表明,该方法能够有效的表征空化强弱程度,具有一定的科学意义及工程价值。After elaborating the techniques of acoustic monitoring and the characteristics of acoustic signal, the cavitation signal is processed with wavelet packet. The characteristic frequency band analysis of wavelet packet is introduced, in which, the cavitation signal characteristics is represented by the characteristic parameters of characteristic frequency band energy and correlation dimension. The cavitations under different working conditions are finally classified according to selected characteristics parameters. The simulation results show that the proposed method can effectively express strength degree of cavitation. It has certain scientific significance and engineering value.

关 键 词：空化空蚀 特征提取 小波包变换 分形关联维数 故障诊断 

分 类 号：TK73[交通运输工程—轮机工程] TM312[动力工程及工程热物理—流体机械及工程]