基于分数阶聚能带分析的微弱故障特征提取研究  

Feature extraction of weak faults based on analysis of fractional energy-gathering band

作　　者：梅检民[1,2] 肖云魁[1] 曾锐利[1] 李枫[1] 任金成[1] 

机构地区：[1]军事交通学院汽车工程系,天津300161 [2]军械工程学院火炮工程系,石家庄050003

出　　处：《振动与冲击》2013年第17期138-144,共7页Journal of Vibration and Shock

基　　金：总后勤部预研项目(AS407C001)

摘　　要：提出了一种分数阶聚能带时频累加谱方法,快速实现长数据的时频分析,突出目标分量,用于提取变速器急加速过程微弱故障特征。根据变速器输入轴转速信号及传动比确定分数阶傅里叶变换(Fractional Fourier Transform,FRFT)最佳阶次,对变速器急加速过程振动信号进行最佳阶次FRFT,根据FRFT模值谱确定聚能带,计算分数阶聚能带时频累加谱,通过对比多组正常和故障数据的分数阶聚能带时频累加谱结果和阶比谱结果,验证该方法的有效性。试验结果表明:根据转速信号能快速、准确确定FRFT最佳阶次;选取聚能带内的FRFT结果进行时频分析,计算量小,分辨率高,分数阶聚能带时频累加谱具有聚焦和局部放大的特点,能很好地突出目标分量,抑噪噪声,是提取变速器急加速过程信号微弱故障特征的有效方法。A fractional energy-gathering band's time-frequency aggregated spectrum (FETFAS) was proposed to achieve the fast time-frequency analysis of long data and extrude target components, and applied to extract the weak fault features of rapid accelerating process of a gearbox. The best order of fractional fourier transformation (FRFT) was ascertained according to a rotating speed signal and a transmission ratio. And the vibration signal of a accelerating process of gearbox was processed using FRFT with that best order, the energy-gathering band was fixed from the modulus spectrum of FRFT, then the FETFAS was computed. The FETFAS and order spectra of many groups of normal and fault signals were compared to verify the effectiveness of FETFAS. The experimental results showed that the method to ascertain the FRFT's best order with a rotating speed signal is fast and correct; the time-frequency analysis of FRFT's results in EGB has a less computing amount and a higher resolution; FETFAS has the characters of focusing and zooming, and it is able to extrude the target components and restrain noise very well, so it is an effective method to extract weak fault features from the signal of a rapid accelerating process of a gearbox.

关 键 词：分数阶傅里叶变换 聚能带 时频累加谱 微弱故障 

分 类 号：TH165+.3[机械工程—机械制造及自动化]