Research On Fault Diagnosis Method Of One-way Valve Under Strong Noise Interference

Pulp pipeline is more and more widely used in industrial transportation.Reciprocating high-pressure diaphragm pump is the most important power equipment for pulp transportation.Therefore,whether the high-pressure diaphragm pump can maintain its normal working state will directly affect the production efficiency of enterprises.Check valve is one of the key components of high-pressure diaphragm pump.Because of its high working frequency and complex mechanism,check valve has become the most frequent failure component of the key components of high-pressure diaphragm pump.During the transportation of pulp in pipeline,uneven particle size of minerals and strong rheological properties of pulp have resulted in unexpected,concurrent,multi-source and uncertain unidirectional valve faults in varying degrees,as well as the emergence of non-stationary and non-linear vibration signals of unidirectional valve.And when the high-pressure diaphragm pump is running,the ambient noise and the noise generated by the components around the check valve cause the vibration signal of the check valve to be disturbed by enormous noise.which makes it difficult for traditional methods to achieve good results in fault detection and diagnosis of the check valve.In view of this problem,this paper has carried out the following research:(1)Aiming at the uncertainty that the delay step and reconstruction dimension of traditional singular value decomposition(SVD)method are selected by experience in the decomposition process and the difficulty of diagnosis of check valve fault caused by strong noise interference.A fault diagnosis method of check valve based on correlation coefficient singular value decomposition(SVD)is proposed.Firstly,the original check valve vibration signal is pre-processed by Empirical Mode Decomposition(EMD),and the intrinsic mode function(IMF)obtained by EMD decomposition is reconstructed.The reconstructed signal is input into the correlation coefficient SVD system.The delay step and reconstruction dimension in the process of singular value decomposition are determined adaptively by correlation coefficient,and then the sieve is carried out.The effective target component is selected,and then the selected component is analyzed by envelope spectrum.The experimental analysis shows that this method can extract the fault characteristics of check valve and has strong practicability;(2)In order to solve the problem that(1)although the noise reduction method can play a good role in reducing the vibration signal of check valve,the useful signal will be eliminated and the traditional stochastic resonance can not achieve a good resonance effect in the strong noise background.A fault detection method of check valve based on stochastic resonance(SR)is proposed.Firstly,the noise reduction algorithm in(1)is used to achieve the purpose of noise reduction,and the delay step and reconstruction dimension in the process of singular value decomposition are optimized by correlation coefficient,then the effective component signal is extracted optimally.The parameters of stochastic resonance are optimized by using ant colony algorithm,and the special spectral energy allocation ability of stochastic resonance is the fault characteristic signal of check valve by using noise energy.Experiments show that this method can extract fault features of check valve very well,and has certain feasibility and validity;(3)Aiming at the problem that the traditional stochastic resonance method is only suitable for small parameter condition in fault diagnosis,a variable step-size stochastic resonance method is proposed to analyze the compression ratio of the vibration signal of check valve under strong noise interference,so as to extract the fault characteristics of check valve under strong noise interference.Firstly,EMD denoising pretreatment is applied to the fault vibration signal,and Intrinsic Mode Function(IMF)components containing fault characteristics are selected to analyze the selected intrinsic mode function components with variable step size stochastic resonance(VSR)to complete the fault diagnosis of check valves.Experiments show that this method can break the constraints of traditional stochastic resonance method which is only suitable for small parameter conditions and is better in strong noise dry condition.The fault feature of check valve is extracted under disturbance;In this paper,the check valve of high-pressure diaphragm pump is taken as the research object.Aiming at the problem that the vibration signal of the check valve is disturbed by strong noise and the fault feature information is difficult to extract,a series of methods for fault detection and diagnosis of the check valve under strong noise interference are proposed,and the fault feature extraction and detection of the check valve under strong noise interference are realized..