Research On Online Monitoring And Fault Diagnosis System Of Mine Main Ventilator

Mine main ventilator is an indispensable safety device in coal mine ventilation,it play s an important role in coal mine safety production, so it’s necessary toconduct an effective online monitoring and fault diagnosis. The existing ventilatoroperation al parameter test method, failure diagnosis method and apparatus havesome flaws. This paper have studie d the test method of ventilator principaloperational parameter s, and combined with the modern intelligent detectiontechnology to achieve the performance test and fault diagnosis of ventilatoroperati onal para meter s, it has a great significance in mine safety production.This paper has studied and analyzed the common fault types and performancecharacteristic of mine main ventilation, then defined the main monitoringparameters that can reflect the ventilator running state and performance, and thenstudied and identified the test method of principal parameter s such as vibration,temperature, humidity and pressure, etc. Through analyze the common fault andcharacteristic frequency of the ventilator, determined to use Wavelet NeuralNetwork to establish the mine main ventilator fault diagnosis model.Construc ted the hardware platform taking computer as the core, complete d theelection of the corresponding transducer, data acquisition card, and design ed theant-aliasing filter and signal conditioning circuit. Taking LabVIEW and Matlab asthe development platform, and develop ed the corresponding monitoring and faultdiagnosis program, monitoring module has achieved the ventilator significantparameters real-time monitoring, data processing and data management, faultdiagnosis module took ventilator vibration signal as the fault diagnosis basis,select db9wavelet as the wavelet bases function to resolve the vibration signal,application of wavelet packet decomposition technique to extracted the fandifferent fault characteristic information from different frequency bands, then took the fault characteristic information as the BP network input vector, theventilator common fault (rotor imbalance, non-centering, pedestal loosen ess, bladefault, surge, etc.) as the output vector, use Matlab neural network toolbox toconstruct three-layer BP neural network and establish the fault intelligentdiagnosis system.The experiment do an accuracy analysis on the analog channel and digitalchannel of the test system, then measured the principal parameter s of4-72-11N02.8A type centrifugal ventilation such as temperature, blast volume androtating speed, etc. Then test the vibration under normal state, pedestal loosen essstate and surge state, take these three kinds of vibration as the neural networkoutput vector, the results show that the network model has a shorter computingtime and a higher credibility.Compared with the original equipment, this subject has develop ed theventilator online monitoring and fault diagnosis system which monitoringparameters are more comprehensive, while the fault information capture is moreaccurate, improve the automation, intelligent degree and efficiency of the minemain ventilator performance monitoring and fault diagnosis, it has a practicalsignificance in reduc ing the coal mine safety accidents.