Study And Application On Performance Analysis With Condition Monitoring And Fault Diagnosis System Of The Mine Major Pump

Pump, which directly affects coal mine production and personnel safety, is thekey equipment of coal mine drainage system. Therefore, real time monitoring andfault diagnosis for pump has an important significance.Through the study of the test principles and methods of mine pump performanceparameters, this dissertation applied PLC and sensor technology to monitor theperformance of coal mine pumps. By the analysis of mechanism and characteristicsquantitative of pump common faults, determine the research methods and ideas ofusing wavelet to extract the feature volume of vibration signals and applying supportvector machines for pump fault diagnosis. This design of pump performance analysisand fault diagnosis system can effectively reduce coal mine pump fault accidents, andprovide on-site parameters as a reference for the scientific management and operationof the pump.Pump Performance analysis is a important content of mine drainage technologymanagement. Expressed pump performance parameter is called pump performanceparameters, for example flow, head, speed, power, efficiency and so on. In this system,the hardware design of pump performance analysis section mainly depends on thesensor and the PLC to complete. The sensor transmits measured signal to the PLCcontroller through module, PLC processes and operates the measured parametersignal to obtain pump performance parameters, and transmits the pump performanceparameters to the host computer, thus the staff in the ground control center canachieve real-time remote monitoring through PC interface.In the part of fault diagnosis of water pump, based on the pump mechanism ofseveral common faults and in-depth analysis of the relationship with the frequenciesof vibration signals, each type of failure in vibration characteristics with a particularfrequency component in the signal,through Wavelet transform for denoising andfeature extraction of vibration signal frequency. In computer implementations, aMatlab data acquisition platform is used to collect characteristic frequency, as sampleswill be collected using grid-based search algorithm for support vector machines tohandle the sample data, achieving for fault diagnosis of water pump. Experimentsshowed that this method of fault diagnosis with high precision and time overhead issmall, therefore for the fault diagnosis of water pump provides practical value.