The Emulsifier Fault Diagnosis System On EMD-sample Entropy And Neural Network Theory

Emulsion explosive is a new type of environmental protection type of explosive.It has the characteristics of strong detonation,good water resistance and so on.And the emulsified explosive production process is simple,high capacity,low production cost,so it has been widely used in our country civil explosive industry.Emulsifier is the core equipment of emulsion explosive continuous,automatic production line.It is a kind of improved rotating machinery equipment,and has the characteristics of high speed rotation.Emulsifier actual operating environment is complex,once faults occur,not only affect the normal operation of the whole production line,it will cause the severe large economic losses,and even the plane crash accident.In order to ensure the safe operation of emulsifying equipment,accurate identification of hidden fault,reduce maintenance costs and improve the utilization rate of equipment,this paper develops a set of fault detection and diagnosis system of Emulsifier,and has been successfully applied in practical production.In this paper,the fault symptoms of these faults are discussed in the view of rotor fault,bearing fault,respectively,and the vibration mechanism of each fault is studied.Firstly,the fault feature extraction method of vibration signal based on sample entropy is proposed.The disadvantage of this method is that the information is limited,and the fault identification is not distinct,so an optimization method of EMD pretreatment sample entropy is put forward.In this method,the vibration signal is decomposed into several IMF components by using EMD firstly,some representative IMF components are selected,and the sample entropy of these components is used as the fault feature.EMD method can dig out the the information that hidden in the signal inside.It effectively overcomes limitation of sample entropy for information acquisition.The results show that the method of EMD combined with sample entropy can not only distinguish different types of fault types,but also can improve the fault tolerance rate of recognition system.The training samples of vibration characteristic parameters were constructed by the vibration history data of the emulsifier operation and fault.Fault feature is used as the input of BP neural network,and the fault recognition is performed by the trained network.The results show that BP neural network can identify the fault types of rolling bearings quickly,and the diagnosiseffect is good.Complete hardware configuration on the basis of the existing equipment of emulsion explosive production line.Based on the industrial control software KingView platform to realize the data exchange of PLC.Through the VB calls MATLAB neural network function to realize the development of the host computer fault diagnosis system.The practical results show that the fault diagnosis system can accurately identify the fault types of the Emulsifier according to the actual data.It has high diagnostic accuracy and good practical application effect.