【摘要】 针对滚动轴承全寿命退化状态难以有效识别的问题,提出了一种基于变分模态分解(VMD)与支持向量机(SVM)相结合的滚动轴承退化状态识别方法。该方法先用包络熵确定VMD的最优分解层数,再根据峭度及相关系数准则选择VMD分解后的敏感本征模态分量(IMF),然后提取敏感IMF分量的时域指标和能量熵构成退化特征向量序列,最后随机抽取不同退化状态下的少量样本输入SVM模型训练,建立退化状态模型库,并用真实数据进行测试。实验结果表明该方法能够准确识别出轴承的退化状态,通过与EMD-SVM、EEMD-SVM模型对比,验证了该方法的优越性。更多还原

【Abstract】 Aiming at problems of rolling bearing’s whole life degradation state being difficult to identify,a Hybrid method based on the combination of the variational mode decomposition(VMD)and Support Vector Machine(SVM)was proposed.With this method,the optimal decomposition layer number of VMD was determined by envelope entropy.The sensitive intrinsic mode function(IMF)component was selected after VMD decomposition according to the kurtosis and correlation coefficient criterion.Then,the time-domain index and energy entropy of the sensitive IMF component were extracted to form the degenerate eigenvector sequences.Finally,a small number of samples from different degraded states by being randomly selected were input into SVM model to be trained,a degenerate state model base was conducted,and it was tested with the real data.The test results showed that the proposed method can be used to accurately identify the degradation state of rolling bearing,and the superiority of the proposed method was verified by comparison with the EMD-SVM,EEMD-SVM model.更多还原