In view of the nonlinear and overlapping characteristics of the fault information of the rotating machinery of petrochemical units,a fault diagnosis method based on the dimensionless and support vector machine(SVM)is proposed.Firstly,the collected vibration signals are analyzed and non-dimensionalized.Then,through feature selection,the dimensionless features with high value and strong sensitivity are selected to reduce the complexity of classification model and improve the speed of algorithm.Finally,the appropriate SVM classification model is selected for classification diagnosis.The fault sensitivity of dimensionless features and the nonlinear classification of SVM are combined for diagnosis and classification.Through the verification by the petrochemical unit fault diagnosis experimental platform,this fault diagnosis method is proved to have a better classification effect than the other classical classification methods,and the classification accuracy can reach 99.1%.
维普
万方数据