Study On Condition Monitoring And Fault Diagnosis Of Emulsion Pump With High Pressure And Large Discharge

Emulsion pump with high pressure and large discharge is a hydraulic power source designed for the supporting equipments in the mega-tonnage comprehensive mechanized coal face of coal industry and is the key equipment to realize greater productivity and higher efficiency of coal face. So the working continuity, stability and reliability of emulsion pump are very important to guarantee the safety in production and to improve the economy of coal industry.Based on the development of automatic devices such as electro-hydraulic unloading valve, backwash filter and electric fluid mixture assembly and the practical necessity and functional requirements to the condition monitoring system of emulsion pump, eight condition parameters are selected as monitored objects and programmable logic controller is selected as the main block in the monitoring scheme. So the condition monitoring system of emulsion pump is developed. The test result shows that the monitoring system of emulsion pump based on programmable logic controller meets the specifications to guarantee the safe operation of equipments.Crank bearing is the key component of crank connecting rod mechanism of emulsion pump and the damageable part during power transmission. The wear fault of crank bearing has been the main fault type of emulsion pump with high pressure and large discharge. Through the analysis of axis eccentricity and minimum oil film thickness of crank bearing under condition of Hydrodynamic Lubrication, the development law of eccentricity and minimum oil film thickness with the variation of radius gap values is generalized, which provides the theory basement of design and maintenance of crank connecting rod mechanism and discloses the fault mechanics and represent feature of crank bearing wear to provide the theory basement to diagnose the wear fault with the vibration signals.Based on the research of crank bearing wear fault mechanics, the fault feature extraction and fault diagnosis method are further studied. Through the power spectrum analysis to vibration signals filtered with wavelet transform in different wear degrees, the variation regularity of information feature and energy of signals is generalized. With the energy analysis to wavelet packet frequency bands, the feature of relative energy of frequency bands can be extracted accurately, which provides the diagnosis of crank bearing wear fault with quantitative foundation. Because the variation of wear degree is gradual and the quantity of fault specimen measured in coal face is finite, the pattern recognition method of fuzzy support vector machine is put forward. Further, the wear fault diagnosis model combining wavelet packet frequency bands energy analysis and fuzzy support vector machine is established. The diagnosis result indicates that the diagnosis model can make full use of fuzzy information of crank bearing wear in different degrees to realize the early diagnosis of wear fault accurately.Pump valve spring fracture is another common typical fault of emulsion pump with high pressure and large discharge which reduces the volumetric efficiency of emulsion pump greatly. According to the solution to the movement differential equation models of pump valve under conditions of normal, separate spring fracture of suction valve and discharge valve, the movement regularity of pump valve can be obtained to disclose the fault mechanics and represent feature of spring fracture to provide the theory foundations for the identification and localization to the spring fracture fault of pump valve with the vibration signals.On the basis of the research of pump valve spring fracture fault mechanics, the singularity analysis method based on continuous wavelet transform is studied and the improved calculation method of Lipschitz exponent based on least median of squares is put forward which is more robust and effective. So the diagnosis rule can be determined. The diagnosis result indicates that this method can realize the correct localization to the singularity points in signal to identify the impulse time under condition of spring fracture and can diagnose the spring fracture fault of pump valve accurately and effectively.