| Rolling bearing is the core components of rotating machinery,and their good working conditions are essential to ensure the safe operation of mechanical equipment.In this paper,in response to the problems of non-stationary,non-linear,fault feature extraction and low classification accuracy during the operation of rolling bearings,a new method based on the combination of variational mode decomposition(VMD)and optimized support vector machine(SVM)is proposed.First of all,for the problem of determining the value of the decomposition layer number K in the variational modal decomposition algorithm,a method based on the center frequency is proposed to determine it.Firstly,the original signal is processed by VMD,and several instinct mode functions(IMF)are obtained,and the center frequency of each IMF under different K values is observed to determine the K value.In order to testify the effect of this method,this paper uses simulation to conduct experimental analysis which based on the empirical mode decomposition(EMD)and variational mode decomposition theories.The results show that the VMD method to determine the parameters through this method can significantly improve the modal aliasing phenomenon and the end effect in the signal decomposition process,and has a good processing effect on the extraction of the fault characteristic frequency of the vibration signal.Secondly,the data of the rolling bearing fault simulation experiment platform of Western Reserve University is used for experimental analysis.The original signal is decomposed into several instinct mode function components after VMD processing,and each component is subjected to envelope demodulation operation,and finally according to the best The envelope spectrum of the IMF can determine the failure of the bearing.Besides,other bearings with different types of failures in the inner ring and outer ring were manufactured by manually,and their vibration signals were collected through an experimental test platform.Similarly,after the signal is processed by VMD,several IMF components are obtained,and the envelope demodulation operation is performed on them,and the fault characteristic frequency of rolling bearing is extracted from them,which verifies the effectiveness of the VMD method in practical applications.Finally,a fault diagnosis method of rolling bearing based on Whale Algorithm(WOA)optimized support vector machine is studied.This paper introduces its concept of the sample entropy,whale algorithm,support vector machine and other theories.In order search penalty factor ‘c’and kernel function parameter ‘g’ in support vector machines,An optimization method based on whale algorithm is proposed.It also uses the rolling bearing data of Western Reserve University in the United States for verification.First,the original signal is subjected to variational modal decomposition to obtain several modal components,and then the sample entropy is extracted from each IMF component to form feature vectors of different fault types,which are used in support vector machines.Training and testing.The whale algorithm is used to select the important parameters [c,g] in the SVM to create a WOA-SVM model for the classification of the rolling bearing state,so as to realize the rolling bearing’s fault diagnosis.Finally,the effectiveness of the model is verified through comparative experiments. |
