Research On Fault Diagnosis System Of Emulsifier Based On Support Vector Machine

The development of explosive performance has brought a great impetus to the development of human production,which greatly reduces the time needed for the development of industrial engineering and has been widely used in mining,road construction and other fields.The development of explosive performance also has brought stricter requirements to the production safety,but in recent years,some serious safety accidents have occurred,which has brought serious damage to property and casualties.With the development of modern technology,the automation level of explosive production line is becoming higher and higher,new ideas and methods are needed for the safety monitoring of emulsion explosive production line.Support vector machine based on accounting method is becoming a hot research direction,which has been widely applied to the research of fault diagnosis.After studying the factors of the faults of emulsion explosive production line,this paper selects the emulsifier which has most faults inear sample data and has unique advantages in the application of fault diagnosis.To sum up,put forward the emulsifier fault diagnostic system based on support vector machine.The main contents of this paper are as follows:(1)According to the factory for water in oil emulsion explosive production line production process principle and procedure and the easy failed emulsifier is paid more attention.Through analyzing the structural mechanism of the emulsifier and studying the type and characteristics of frequent faults,the vibration signal of emulsifier is selected as the fault detection signal.(2)After comparing several method of machinery fault diagnosis,combined with emulsifier fault features and support vector machine to the demand of fault samples is low characteristics is proposed based on support vector machine emulsifier fault diagnostic system of the fault for detection and diagnosis.Select the local mean decomposition of fault signal and calculate the characteristic value.And the mature software Libsvm is selected as detection software.(3)According to the existing software and hardware of the emulsion explosive production line,the hardware and software platform of the system is redesigned.About the hardware platform,the selection of vibration sensor,signal transmitter,host computer and t PLC is introduced and analyzed in detail.The software platformintegrates Matlab program,Kingview program and PLC program to realize the communication between hardware and software by using VB program.(4)Based on the characteristics of the vibration of the emulsifier,the appropriate fault feature data is selected as the research object.The selection of the SVM kernel function and parameters,and the establishment of the model are completed.Simulation results show the effectiveness of the diagnostic system.Finally,the research work is summarized and the further research direction is pointed out.