Fault Diagnosis Of High Pressure Diaphragm Pump Check Valve Based On CEEMD And LSSVM

The high-pressure diaphragm pumps servers as the core power equipment of mineral slurry pipeline.It is essential to ensure its normal working condition.Check valve is one of the core components of the high pressure diaphragm pump,it is more vulnerable to failures since it works in high pressure,abrasion,corrosion slurry environment.In order to ensure safe and efficient operation of the high pressure diaphragm pump,fault diagnosis of check valve is particularly important.However,it is difficult to collect the vibration signal of the check valve.The vibration signal of the check valve is nonlinear and non-stationary,which brings great difficulty to the signal processing and fault diagnosis.The running state of the check valve is related to the conveying pressure and the rheological characteristics of the slurry,and the fault is sudden,which makes the fault diagnosis more difficult.Therefore,it is of great significance to carry out the operation state monitoring and fault diagnosis.In this paper,the vibration signal of the check valve is taken as the research object to study signal processing,feature extraction,classification and fault diagnosis.The research contents of this paper are as follows:(1)This paper presents an improved wavelet threshold denoising method base on hard thresholding method and soft threshold method.The proposed method can effectively eliminate the Gibbs effect produced by the hard threshold method at the discontinuous points of the signal,and can solve the problem that the error between the reconstructed signal and the original signal is large in Soft-threshold method.The simulation results show that the proposed method can achieve good noise reduction effect,and can be used to realize the noise reduction of check valve vibration signal.(2)This paper presents a fault diagnosis method of check valve based on the Complementary Ensemble Empirical Modal Decomposition(CEEMD)and Intrinsic Mode Function(IMF)selection and Least Squares Support Vector Machine(LSSVM)classification model.The simulation experiment compares the performance of three kinds of signal decomposition methods.Because the CEEMD method can effectively solve the problem of mode mixing and zero failure in the Empirical Modal Decomposition(EMD)method,the CEEMD method is used to analyze the vibration signal of the check valve.At first,the IMF components were analyzed by correlation and energy comparison,and then Singular Value Decomposition(SVD)is used to extract the features of the filtered IMF components.Finally,the feature vectors are input into the LSSVM model to diagnose the faults.Experiments show that LSSVM has higher recognition rate than Support Vector Machine(SVM).(3)This paper presents a fault diagnosis method of check valve based on Local Mean Decomposition(LMD)and Particle Swarm Optimization and Least Squares Support Vector Machine(PSO-LSSVM)classification model.First of all,the LMD method is used to decompose the vibration signal of the check valve,and then SVD is used to extract the features of the PF components which removing residual component.Finally the feature vectors are input into the PSO-LSSVM model to diagnose the faults.Experiments show that PSO not only has strong ability of searching,and can avoid the occurrence of local optimization.(4)Using mixed programming method of C#and MATLAB to complete the development of check valve fault diagnosis system.In this paper,the vibration signal acquisition of the check valve of high pressure diaphragm pump is completed,and the fault diagnosis system of check valve is tested.Aiming at the problems of noise reduction,feature extraction and fault diagnosis in nonlinear and non-stationary signals,a series of effective noise reduction methods and fault diagnosis methods are proposed in this paper.The simulation results show that the proposed method is effective and feasible.This method are applied to the analysis of the vibration signal of the check valve of the high pressure diaphragm pump.it makes the fault diagnosis of check valve with higher accuracy and convenience.The method provides a theoretical basis and reference for the vibration signal analysis and fault diagnosis of mechanical equipment while reducing the economic loss caused by check valve failure.