Study Of Vibration Analysis And Malfunction Diagnostic For Mine Ventilator

Coal main fan is one of the most important equipment in coal mine, as the mine lung , its chief task is to supply air for the whole mine and its running well or not effects on the safety production directly to the coal mine. So it is necessary to do some fault diagnosis and earning worming research, that can be of significance to both safety production of coal mine and maintenance of devices.This paper has summed up the developing situation, research and trend of development of coal mine fan's fault diagnosis at home and abroad. On the base of research on the coal mine fan's vibration mechanism and Analytical method, theoretical analysis, numerical simulation and experimental research are all used in intensive study based on Wavelet Neural Network and evidence theory for main fan fault diagnosis and predictive failure analysis. Software for main fan online monitoring and fault diagnosis has been developed that makes its online fault diagnosis come true. Test results proved that this method has optimized the classical neural network algorithm, multiplexed the advantage both Wavelet Neural Network and evidence theory, improved the speed and precision in mine fan fault diagnosis.The primary works and research findings in this paper as follows:①The analysis method of vibration based on the tower model is attained by researching the vibration mechanism of coal mine fan. It is a theoretic reference for engineers majoring on the fault diagnosis of rotary machine.②Besides the research of the fault reasons and the studies of the time waveform and the spectral characteristics when the various faults are happening, some fault diagnosis is also done according to the FFT spectral characteristics analysis method.③The mapping form characteristic space to fault space is acquired by taking advantage of the artificial neural network nonlinear mapping method and forecasting model of wavelet network structure is established. The development trend of ventilating fan vibration intensity may be confirmed with the help of the model, that can provide reference for grasping the fault and maintaining the system in advance.④The fault diagnosis method that integrates neural network and proof theory is achieved. This method fully takes advantage of their merits and introduces in detail the application of the model in fault diagnosis of mine main fan. It is proved that this method is very useful in practice.⑤On the basis of KingView, through Matlab and VC++ mixing programming technology, main ventilating fan online monitor and fault diagnosis software is developed with the benefit of information fusion. In addition, the system fault diagnosis early warning platform is also achieved.