摘要
滚动轴承在发生故障时,故障振动信号具有非稳定性、非线性的特点,难以对其中的故障特征进行提取,导致轴承故障诊断的识别率较低。为了提高滚动轴承故障分类的准确率,提出了一种基于集合经验模态分解法(Ensemble Em pirical Mode De com pos ition, EEMD)与长短时记忆(Long Short Te rm Me m ory, LSTM)神经网络相结合的滚动轴承故障识别的方法。首先采用EEMD算法将目标振动信号分解成若干个本征模态函数(Intrinsic Mode Function, IMF)分量。然后利用主成分分析法(Principal Component Analysis, PCA)对IMF分量进行降维,选取含有主要故障特征信号的分量。最后计算IMF主成分分量占各自总能量的比例,并将能量比所组成的特征向量作为LSTM神经网络的输入参数进行故障识别。将识别的结果与不同的故障诊断模型所得的结果进行对比分析,仿真结果表明文中所用的方法在轴承故障诊断中准确率更高。
When rolling bearing is in fault,the fault vibration signals are unstable and nonlinear,and it is difficult to extract the fault features,which leads to low recognition rate of bearing fault diagnosis.In order to improve the accuracy of rolling bearing fault classification,a rolling bearing fault identification method based on Ensemble Empirical Mode Decomposition(EEMD)and Long Short Term Memory(LSTM)neural network is proposed.Firstly,the target vibration signal is decomposed into several Intrinsic Mode Function(IMF)components using EEMD algorithm.Then,Principal Component Analysis(PCA)is used to reduce the dimensions of IMF components,and the components containing major fault characteristic signals are selected.Finally,the proportion of IMF principal components in the total energy is calculated and the eigenvector composed of the energy ratio is used as the input parameters of the LSTM neural network for fault identification.The results of identification are compared with the results of different fault diagnosis models,and the simulation results show that the method used in this paper is more accurate in bearing fault diagnosis.
作者
杨淑洁
周杨
YANG Shujie;ZHOU Yang(College of Ocean Engineering Equipment,Zhejiang Ocean University,Zhoushan 316022,China)
出处
《机械工程师》
2021年第11期28-33,共6页
Mechanical Engineer
