Fault Diagnosis Research Of Mine Gearbox Based On Multiwavelets Transform

China is the first coal producing country,and the coal industry is an important basic industry in China.At present,coal mine safety has become the main problem that restricts the development of the coal industry.To prevent and control the occurrence of serious accidents in coal mines,and to promote the safety and stability of the coal mine has become a major problem to be solved at the national and government levels.Coal mine safety is the most important work of national industrial safety.Large coal mine machinery(such as hoist,belt conveyor,coal shearer)power transmission gearbox working environment is bad,the probability of failure is very large.Its running state directly affects the reliability and safety of the operation of these large mechanical equipment,once the failure occurs,it will cause huge economic losses to coal mine safety production,and even casualties.Therefore,it is very necessary to carry on the condition monitoring and fault diagnosis of the coal mine transmission gearbox.The advanced mechanical fault diagnosis method is applied to the fault diagnosis of mine gearbox,which is of great significance to ensure the safe production of coal mine and to avoid safety accident.Wavelet transform has become a very wide range of non-stationary signal processing tool,has been widely used in separation of fault feature frequency,extraction of weak signal,and early fault diagnosis of rotating machinery.However,in the mechanical fault diagnosis,the wavelet basis function that wavelet transform selected is determined,which can not be changed characteristics according to the signal characteristics,it is difficult to achieve the best match with the fault characteristic waveform.Second generation wavelet can construct wavelet basis function in the time domain,by changing the predictor and updater coefficient,construct wavelet function matching fault feature,make up for the shortcomings of classical wavelet.But whether it is classic wavelet or second generation wavelet,they only have a wavelet basis function,usually only better matching single type fault feature signal,at the same time it is difficult to identify different types fault feature.Daubechies proved that the single wavelet can not satisfy the orthogonality,symmetry,compact support and high vanishing moments and other excellent properties,which are very important in signal processing.This is a prominent problem in fault diagnosis application field which classic wavelet and second generation wavelet faced,Therefore,in order to solve the above problems,scholars at home and abroad focus the research on multiwavelets.Fault feature extraction technology is the key technology in mechanical equipment condition monitoring and fault diagnosis,with the advanced theories and methods to enrich and improve the mechanical fault diagnosis technology is an important content of mechanical fault diagnosis.In this paper,solve the large coal mine mechanical equipment fault diagnosis as the engineering application requirements,using modern signal processing technology of multiwavelets theory as a tool,Taking the gearbox of the large scale mechanical equipment in coal mine as the research object,carried out the fault diagnosis.Gearbox vibration signal is complex,how to effectively extract the implicit fault feature information from the gearbox vibration signal,which for the gearbox condition monitoring and fault diagnosis is very important,and it is the key.In this paper,new methods of gearbox fault feature extraction based on multiwavelets transform are studied,and they are successfully applied to the fault feature extraction and fault diagnosis of the test gearbox and mine gearbox.The main research contents are:(1)The research background and significance of the research are described.Analyzes the research status quo of gearbox fault diagnosis at home and abroad,as well as multiwavelets theory,produces the research question of multiwavelets transform in large coal mine mechanical equipment gearbox fault diagnosis,establishes the research route and the main research content of this article.(2)Research on coal mine hoist gearbox fault diagnosis based on neighboring coefficients of translation-invariant multiwaveletsExplained traditional multiwavelets threshold denoising method,study the basic principle of translation-invariant multiwavelets denoising using neighboring coefficients,using the simulation signal analysis and test gearbox vibration signal de-noising experiment,compared to single wavelet,multiwavelets,translation-invariant multiwavelets denoising effect of hard threshold,soft threshold and neighboring coefficients.Through the comparison of correlation coefficients of denoised signal with the original signal,the translation-invariant multiwavelets denoising using neighboring coefficients has the best denoising effect.Finally translation-invariant multiwavelets denoising using neighboring coefficients is applied to fault diagnosis of coal mine hoist gearbox,The vibration signal of the gearbox is denoised effectively,extract the fault feature of gearbox accurately,judge the fault location of the gearbox,provides an effective method for large coal mine machinery equipment gearbox vibration signal denoising.(3)Coal mine shearer transmission gearbox fault diagnosis based on multiwavelets decomposition combining maximum correlation kurtosis deconvolutionStudy on the band characteristics of multiwavelets decomposition,the band selecting criterion of kurtosis criterion and correlation coefficient criterion for fault analysis,and the basic theory of maximum correlation kurtosis deconvolution,we propose a fault diagnosis method of multiwavelets decomposition combining with maximum correlation kurtosis deconvolution.make the experiment analysis of test gearbox broken teeth and local broken tooth fault,gearbox vibration signal for multiwavelets decomposition,selecting the frequency band that correlation coefficient and kurtosis is larger for maximum correlation kurtosis deconvolution denoising analysis,the weak shock fault feature has been strengthened,then carry on envelope spectrum analysis after maximum correlation kurtosis deconvolution denoised,successfully extracted gearbox fault feature frequency.Finally,the method is applied to the fault diagnosis of coal mine shearer transmission system gearbox,and play their respective advantages,effectively extract the fault feature information,and realize the accurate diagnosis of the gearbox.(4)Coal mine belt conveyor gearbox fault diagnosis based on multiwavelets decomposition and constrained independent component analysisIt introduces independent component analysis and constrained independent component analysis algorithm,analyzes their shortcoming and advantage,study the ICA model of multi-channel measurement signal containing source noise,combined with the advantages of multiwavelets transform,proposes a gearbox fault feature extraction method based on multiwavelets transform and constrained independent component analysis.Through the fault feature extraction test of testbed gearbox,according to the kurtosis criterion,Select the kurtosis larger component compose mixed signal from multichannel components of multiwavelets decomposition of gearbox vibration signal.Take gearbox meshing frequency as a priori information to establish a reference signal,using constrained independent component analysis algorithm successfully extract gearbox fault feature frequency.Finally,the proposed method is applied to the fault diagnosis of coal mine belt conveyor gearbox,judge the fault of the gearbox,and provide a new method for the fault feature extraction of gear box.Experimental analysis and coal mine engineering application results verified the effectiveness of the proposed method.(5)Application of redundant multiwavelets packet spectral kurtosis method in mine gearbox fault feature extraction.Study the basic principle of spectral kurtosis and fast algorithm – Kurtogram.Describes the decomposition algorithm of multiwavelets packet,and the algorithm of redundant multiwavelets transform,apply redundant multiwavelets packet decomposition to replace the traditional spectral kurtosis filter,proposed redundant multiwavelets packet spectral kurtosis method.With an fault diagnosis example of test and mine gearbox to validate the method,use redundant multiwavelets packet spectral kurtosis algorithm to decompose the vibration signal of the gearbox,obtain the optimal filtering frequency band signal,and then demodulates analysis,successfully extract fault characteristic frequency of gearbox.The analysis results were compared with the traditional spectral kurtosis method,redundant multiwavelets packet spectral kurtosis method can accurately select the optimal filtering frequency band,avoid noise interference,fault feature extraction is better than the traditional spectral kurtosis method.Experimental analysis and coal belt conveyor gearbox engineering application proved that the proposed redundant multiwavelets packet spectral kurtosis method is an effective fault feature extraction method.