关联维数和基于内禀模态函数的AR模型在滚动轴承故障诊断中的应用

提出一种基于内禀模态函数(Intrinsic Mode Function,IMF)、自回归(Auto-Regressive,AR)模型和关联维数的滚动轴承故障诊断方法.该方法首先采用经验模态分解(Empirical Mode Decomposition,EMD)将滚动轴承振动信号分解成若干个IMF,然后对包含主要故障信息的IMF分量建立AR模型,计算AR模型自回归参数的关联维数,并以关联维数作为特征向量输入神经网络分类器,最后通过网络的输出结果来识别轴承的工作状态和故障类型.对实验数据的分析结果表明,该方法能有效地应用于滚动轴承的故障诊断.

Application of correlation dimension and IMF based AR model to fault diagnosis for roller bearing

A fault diagnosis approach for roller bearing based on IMF(Intrinsic Mode Function),AR(Auto-Regressive)model and correlation dimension is proposed.First of all,vibration signals of roller bearing are decomposed into a finite number of stationary IMFs by EMD(Empirical Mode Decomposition)method.Then the AR model of IMF component that contain key fault information is established.Finally,the correlation dimensions of auto-regressive parameters in AR model are calculated and served as input vectors of neural network.Thus the fault patterns of roller bearing can be identified by the output of the neural network.The experimental results show that the proposed approach can be used to identify roller bearing fault patterns accurately and effectively.