基于AFD算法的滚动轴承故障诊断方法  

Fault Diagnosis Method of Rolling Bearing Based on AFD Algo rithm

作　　者：梁瑜[1,2] 贾利民[1,2] 蔡国强[1,2] 刘金朝[3] 

机构地区：[1]北京交通大学交通运输学院,北京100044 [2]北京交通大学轨道交通安全与控制国家重点实验室,北京100044 [3]中国铁道科学研究院基础设施检测中心,北京100081

出　　处：《中国铁道科学》2013年第1期95-100,共6页China Railway Science

基　　金：国家"八六三"计划项目(2010AA1100830006);国家科技支撑计划项目(2009BAG11B02)

摘　　要：基于自适应傅里叶分解(AFD)算法,将滚动轴承的振动信号分解为一系列单一分量信号并计算它们的峭度;将峭度由大到小顺序排列,自适应寻找峭度趋于稳定的拐点,对拐点前的单一分量信号求和并取包络作共振解调;根据解调得到的频谱判断滚动轴承是否发生故障及发生故障的部位。以N205EM型滚动轴承为例进行试验验证,结果表明:在不预先确定滤波频带,不出现无物理意义的"负频"情形下,能够准确有效地提取出比传统共振解调方法有更好频谱特征的滚动轴承故障信息,从而有效地诊断出滚动轴承的故障。Adaptive Fourier decomposition （AFD） algorithm decomposes the vibration signal of rolling bearing into a series of mono-components, and the kurtosis of each mono-eomponent is calculated. The kurtosis is arranged in descending order. The inflection point of kurtosis becoming stable is adaptively sought out and the corresponding mono-component signals before inflection point are summed up, then the envelope is taken as the resonance demodulation. According to the frequency spectrum obtained from de- modulation, whether rolling bearing has fault is judged and the fault location is determined. Taking N205EM-type rolling hearing for example, the experiment results indicate that the proposed method is ac- curate and effective in extracting the fault iiaformation of roiling bearing without presetting filter frequency band and the absence of ＂negative frequency＂ with no physical meaning. The spectrum characteristics bet- ter than traditional resonance demodulation are obtained. The proposed method is effective in diagnosing the fault of rolling bearing.

关 键 词：滚动轴承 故障诊断 自适应傅里叶分解 单一分量信号 峭度 共振解调 

分 类 号：U260.331.2[机械工程—车辆工程]