Health Management Methods Of Rolling Element Bearing

Rolling element bearings are widely used in rotating machinery,and play a vital role in the modern industrial production.The accidental fault of rolling bearings in mechanical equipment may cause huge economic losses and even casualties.Therefore,it is of great significance to study the health management of rolling bearing,to realize the initial fault diagnosis,performance degradation stages division and remaining useful life prediction.It is beneficial to the timely maintenance of equipment,the formulation of maintenance strategy and the prediction of remaining useful life in the current state.The run-to-failure rolling bearing vibration signal is taken as the research object,and the proposed health management methods are combined with the modern signal processing technology to study the initial fault diagnosis,performance degradation stage division and remaining useful life prediction of rolling bearing.The main research contents of the paper are as follows:(1)The method for initial fault diagnosis of rolling bearings is proposed based on the coarse-to-fine strategy.The initial fault signal of rolling bearings is easily overwhelmed by the strong background noise,and the different characteristic parameters have different sensitivities to the early fault.The Intrinsic Mode Functions(IMFs)of the original vibration signal are obtained by using Variational Mode Decomposition(VMD).The singular values of IMFs are obtained by using Singular Value Decomposition(SVD).The singular values and the time domain characteristic parameters sensitive to initial faults are fused into MD2.VMD,Autogram,Multipoint Optimal Minimum Entropy Deconvolution Adjusted(MOMEDA)and envelope analysis are used to diagnose the fault type of rolling bearing.The initial fault diagnosis of rolling bearings is realized according to the proposed 3? principle of ?MD2 and the backtracking algorithm.The validity and superiority of the proposed method are verified through the experimental results and discussions.(2)The method of Health Indicator(HI)construction based on MD-CUMSUM is proposed for the research of rolling bearing degradation stages division.The performance degradation stages of rolling bearing are difficult to be accurately divided,and the performance evaluation indicator is fluctuant and nonmonotonic.VMD-SVD is used to extract the time-frequency domain features of vibration signal.The time domain characteristics related to performance degradation combined with the time-frequency characteristics are fused into MD1.CUMSUM is used to process MD1,and the HI monotonically increasing with bearing wear is obtained.The initial fault of rolling bearing is determined according to the proposed coarse-to-fine diagnosis strategy.The performance degradation stages of rolling bearing are divided based on the initial fault sample and the reasonable thresholds of HI.The effectiveness and superiority of the proposed method in fault stages division are verified through the experimental results and discussions.(3)The method for remaining useful life prediction of rolling bearings is proposed based on the risk assessment and improved training set.The initial fault and failure time of rolling element bearing should be weighed between the sensitivity and false alarm rate.In order to obtain a better result of remaining useful life prediction,the time-frequency domain features extracted by VMD-SVD and the optimal time domain features are fused into MDs.The monotonically increasing health indicator is obtained through MD-CUMSUM.The initial fault time and failure time of rolling bearings are weighted by using the proposed risk assessment method.The proposed method of improved training set S and the optimized Support Vector Regression(SVR)by genetic algorithm is utilized to achieve the remaining useful life prediction of rolling element bearings.The accuracy and superiority of the proposed method in the prediction of remaining service life are verified by the experimental results and comparisons.