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

As one of the key equipments which is very easy to broken down in the production line of emulsion explosive,It is highly necessary to design a reasonable and efficient fault diagnosis system of emulsifier.Aiming at the non stationary and small sample of the fault vibration signal of the emulsifier,in this paper,a method based on the EMD method is introduced to obtain the fault features of the emulsion.On this basis,a fault diagnosis system based EMD and SVM is designed,and verification and analysis are carried out in the laboratory environment.This paper mainly includes:(1)Firstly,this paper introduces the whole process of emulsion explosive production,then takes the AE-HLC-III emulsifier as research object according to the selection of emulsifier,and introduces the basic structure and advantages of the emulsifier in detail.On this basis,several typical fault types and vibration mechanism of the emulsifier are studied.(2)Extract the fault characteristics of the emulsifier by using EMD method.The collected fault signals of the emulsifier often have non stability,which must be extracted.In this paper,original signal is decomposed into a series of intrinsic mode function(IMF)by using the EMD method,then calculate the energy of every IMF component to constitute a feature vector of emulsifier,at last this feature extraction method is proved the feasibility.(3)Use genetic algorithm to optimize the kernel parameters and penalty factors of support vector machine.The classification performance of SVM is mainly dependent on the penalty factor and kernel parameter.The genetic algorithm has the convergence,and it can find the global optimal solution in the search process in high probability,therefore,this paper uses genetic algorithm to optimize the parameters of SVM,and the diagnosis results show that the SVM has good classification ability.(4)A fault diagnosis algorithm based on EMD and SVM is proposed.In fact,the fault characteristics of the emulsifier are often small samples due to various reasons.SVM is a theory based on small sample learning,which can establish an effective fault diagnosis model according to the limited fault samples of the emulsifier.The EMD methods can extract the weak fault signal in the noise,and the experimentalresults show that the combination of the two methods has achieved good results.(5)The overall design of the emulsifier fault diagnosis system is put forward,and the hardware and software design of the emulsifier are realized.In the aspect of hardware,the principle of each hardware design is introduced.According to the requirement of emulsion explosive production,the selection equipment is conducted.In the aspect of software,the whole function as well as the operation interface of the software system is introduced.Finally,the effectiveness of the system is verified by the failure of the rotor in the emulsifier.The results show that the system can effectively detect the faults of the emulsifier.