LMD能量矩和变量预测模型模式识别在轴承故障智能诊断中的应用

LMD energy moment and variable predictive model based class discriminate and their application in intelligent fault diagnosis of roller bearing

变量预测模型的模式识别方法(Variable predictive model based class discriminate,VPMCD)是一种利用特征值相互内在关系进行模式识别的新方法。论文提出了基于局部均值分解LMD(Local mean decomposition,LMD)能量矩概念,并针对轴承故障振动信号特征值的相互内在联系,将LMD能量矩与变量预测模型模式识别相结合,提出了一种轴承故障智能诊断新方法。首先利用LMD方法将复杂非平稳的原始信号分解为若干PF(Product function,PF)分量;然后利用相关分析剔除LMD方法中的虚假PF分量,并提取真实PF分量能量矩组成特征向量来有效地表达故障信息;最后采用VPMCD方法进行轴承故障诊断。通过仿真信号验证了PF能量矩比PF能量更能反映非平稳信号本质特征。轴承故障诊断实验结果表明,论文提出的方法能有效地应用于小样本多分类轴承故障智能诊断。

Variable predictive model based class discriminate （VPMCD） is a new pattern recognition approach,which takes full advantage of the inhere relation between the feature value.The conception of LMD energy moment was presented in this paper.Aimed at the inhere relation between the feature value of roller bearing,combing the LMD energy moment and VPMCD,a novel intelligent fault diagnosis method was proposed.Firstly,the complicated non-stationary original vibration signal was decomposed into a set of product function components.Secondly,correlation analysis method was used to remove pseudo-components and the energy moment of real PF components with signal feature was extracted as eigenvector to express the fault information adequately.Lastly,VPMCD was served as the approach of pattern recognition to identify roller bearing fault type.The simulation results demonstrate the energy moment of PF components can reflect essential feature of non-stationary signal.The analysis results from practical roller bearings fault vibration signal show that the proposed method can be applied to small sample multiple classification intelligent fault diagnosis of roller bearing effectively.