基于小波奇异性理论的水轮机空化检测  

TURBINE CAVITATION TESTING BASED ON WAVELET SINGULARITY DETECTION

作　　者：蒲中奇[1] 张伟[1] 施克仁[1] 张俊华[1] 吴玉林[2] 

机构地区：[1]清华大学机械工程系无损检测实验室,北京100084 [2]清华大学热能工程系,北京100084

出　　处：《振动与冲击》2005年第5期74-76,共3页Journal of Vibration and Shock

基　　金：国家自然科学基金项目(50176022)

摘　　要：空化噪声谱分析中常采用的功率谱分析在研究空化噪声这种具有很强突变性的信号时具有明显的缺陷,文中提出了一种基于小波奇异性理论的水轮机空化检测方法,并提出了小波基选择方案、突变点检测条件和最佳检测阈值。进行了混流式水轮机的模型转轮试验,观测了模型转轮额定工况下的涡带形态和空化发展情况,并采集存储了大量空化噪声数据。对照观测结果分析实际的检测数据,结果表明该方法的有效性,并很好的检测出空化初生和空化形态转变。Power Spectrum analysis which is usually used in spectrum analysis of cavitation noise is obviously unfit for investigation on the cavitation noise signal due to its transient character. Turbine cavitation testing based on wavelet singularity detection is presented and mother wavelet option methods, abrupt change points detecting conditions and best testing threshold are apso proposed. Model experiment of a Kaplan model turbine is carried out. Pattern of vortex in draft tube is observed and lots of cavitation data are sampled and stored. The real cavitation noise data are analyzed by contrast with the results of observation and it is proved that the methods are valid and incipient cavitation and change of cavitation pattern can both be detected defectively.

关 键 词：水轮机空化 空化噪声 小波分析 奇异性检测 检测阈值 模型试验 

分 类 号：O427.5[理学—声学]