Application Research Of HHT Time-Frequency Analysis Method In Fan Fault Diagnosis

Mine Ventilators play an important role in the safe production of coal mining industy.If the fan fails,even a brief wind stop may endanger-the safety of the underground workers.Therefore,the in-depth study of fan fault diagnosis technology has certain practical significance.In this paper,the vibration signal of the main fan of the mine is taken as the research object.After summarizing the development status of the fan fault diagnosis technology at home and abroad,a fault diagnosis method of the fan based on Hilbert-Huang transform(HHT)time-frequency analysis and fuzzy neural network is proposed.Firstly,the short-time Fourier transform,Wigner-Ville distribution and Hilbert-Huang transform are compared,and then the Hilbert-Huang transform with more adaptive and local characteristics analysis is selected as the time-frequency analysis method.Secondly,the improved wavelet transform is compared with three different threshold function denoising methods.Finally,the noise reduction method combined with lifting wavelet and semi-soft threshold denoising method is selected,to filter and denoise the vibration signal.By optimizing the modal aliasing phenomenon in empirical mode decomposition(EMD),the time-frequency-energy of the vibration signal is obtained by using the optimized set empirical mode decomposition(EEMD)and Hilbert spectral analysis(HAS).Extracting different energy values in each IMF component frequency band to form a feature vector;Finally,the eigenvector is used as the input of the fuzzy neural network combining the fuzzy system and the neural network,which makes the identification of the fan fault more accurate.Based on the above research results,the time-frequency feature quantity obtained by HHT analysis is used to diagnose and identify the fault type using fuzzy neural network.The diagnosis results are in good agreement with the actual situation.The established diagnostic model has certain reference value on the subsequent system research.