Research On Solenoid Valve Fault Diagnosis

With the increasing demand for high-performance electro-hydraulic control system,the solenoid valve as a vital component of electro-hydraulic control equipment,is widely used in the shipbuilding industry,petrochemical,national defense industry,industrial equipment and other important areas,because of its simple structure,low cost,low-pollution,fast response,etc..Due to long-term wear,corrosion,electrolysis and other factors,the solenoid valve are prone to failure,which will seriously affect the performance of the control system and even lead to system downtime.Therefore,this thesis conducts a in-depth study on the common fault diagnosis method of solenoid valve.In this thesis,the mathematical model of the solenoid valve is established based on the working principle,and the corresponding simulation platform is built on MATLAB / SIMULINK simulation software.Secondly,The feasibility of the model is verified by analyzing the driving current of the solenoid valve and the displacement of the spool,which is detected from the simulantion platform.Then,the fault mechanism of several common solenoid valve fault,such as springbroken,spring superfluence,spoolstuck,spool slight jamming and so on,is analysized in detail,the corresponding simulation is carried out on the platform,and the current waveform of the solenoid valve is collected.What's more,the current conversion rate is decomposed into three layers by the method of wavelet packet transform technique and energy moment technique,and the effective fault feature vector is extracted from it.After analyzing the fault feature vector,the corresponding BP neural network is designed to classify and identify the fault vector.Finally,this thesis chooses the 4WE6E6 X / EG24N9K4 type solenoid valve,designs the STM32-based solenoid valve fault diagnosis physical experiment platform,further validates the simulation model,and valve fault is carried out,the data is collected,and the feature vector is extracted.The thesis designs the off-line fault diagnosis based on BP neural network and on-line fault diagnosis based on association rules,both of which achieve high fault diagnosis accuracy,which further validates the effectiveness of the solenoid valve model and the fault diagnosis method.Therefore,this subject has laid a foundation for the realization of the fault diagnosis of solenoid valve.