The Research Of Emulsifier Fault Diagnosis System Based On EWT And SVM

The emulsifier is an important equipment in the emulsion explosive production line,which is prone to failure due to its harsh working environment and large working intensity.Therefore,it is very meaningful to distinguish the type of failure of the emulsifier in a timely and effective manner.In this paper,the emulsifier bearing was used as the research object,and combined with Empirical Wavelet Transform(EWT)and Support Vector Machine(SVM).It is applied to the fault diagnosis system of emulsifier.In this paper,the normal state,the inner ring fault,the outer ring fault and the roller fault were studied.Firstly,the vibration signals of the four states of the emulsifier bearing were collected separately,and the collected vibration was proposed by the EWT method.The signal was subjected to feature extraction and a feature vector was constructed therefrom.The feature vector was input as sample data into the SVM to train the model,and the classification and recognition of the emulsifier bearing state was realized.In addition,in order to improve the accuracy of classification and recognition,the two common SVM multi-classification methods were compared and analyzed,and the support vector machine model was determined.The particle swarm optimization algorithm was proposed to optimize the parameters of the SVM model.The experimental verification showed that the recognition ability of the optimized support vector machine model was significantly improved,and the EWT and SVM emulsifier fault diagnosis methods were completely feasible.Finally,an emulsifier fault diagnosis system was designed.In terms of hardware selection,hardware equipment is selected in combination with the actual production requirements of emulsion explosives.In software,the diagnostic functions and operation interface were designed in combination with Matlab and Visual Basic.The effectiveness of the system was verified by the inner ring fault of the emulsifier bearing,which achieves the ideal effect and proves the effectiveness of the system.