Fault Feature Extraction Of Gearbox On Variable Rotating Speed Working Conditions Based On Multichannel Multicomponent Decomposition

As the main transmission and transmission device of power mechanical equipment,the gearbox plays an important role in the stable and safe operation of power machinery and equipment.Therefore,it is of great practical significance to monitor and identify the operating state of the gearbox,of which the extraction of the characteristic quantity that can reflect the operating state of the gearbox is the key.However,the gearbox operating conditions are changeable and excitated,resulting in its vibration signal having the characteristics of multicomponent non-stationary,non-linear,and low signal-to-noise ratio,and will show complex features such as strong time change,close proximity or overlap in the time frequency domain,which will affect the clear and accurate extraction of gearbox fault characteristics.Therefore,how to use effective signal processing and analysis methods to extract the state characteristic information of the gearbox from the multicomponent signal with complex time-frequency characteristics has always been a hot topic in research.The traditional time-frequency analysis method based on the characteristics of the timefrequency scale has certain limitations to different degrees when extracting fault characteristics from the vibration signal of the gearbox with the above-described complex time-frequency distribution characteristics.Based on this,this paper introduces a new Multichannel Multicomponent Decomposition(MMD),and on the basis of solving some of its theoretical defects,it is applied to the fault feature extraction of gearbox under variable speed conditions,and the following research work is mainly carried out:(1)The theory of multichannel multicomponent decomposition method is expounded and analyzed,and the Multichannel Multicomponent Decomposition method is summarized that it has the characteristics of adaptability,orthogonality,and sparseness of decomposition results,and the relationship between the heuristic algorithm adaptability function and the multicomponent quantity search is given,which is the decomposition of MMD algorithm using different optimization algorithms Efficiency in doing research preparation work.According to the characteristics of the vibration signal of the gearbox,the decomposition range of the MMD method is investigated by the simulation signal of different time-frequency domain support characteristics The classical decomposition methods based on different decomposition methods are compared and analyzed,and the analysis shows that the decomposition effect of MMD on adjacent signals and time-frequency overlapping signals is significantly better than that of EMD and VMD Two methods of decomposition.(2)Aiming at the problem that MMD has low decomposition efficiency for high-sample signals,the logarithmic window time-frequency energy criterion is introduced into the MMD decomposition method,and the logarithmic window time-frequency energy criterion can be folded to select the window length,improve the signal time-frequency aggregation,make the time-frequency measurement calculation accuracy higher,the calculation efficiency faster,and improve the decomposition efficiency of the MMD method.(3)Aiming at the problem of poor noise immunity of the number of components of the MMD method,the strongest eigenvalue component estimation method is analyzed as MMD The method is a defect of the method of separation estimation method,and the multichannel time-frequency slice accumulation algorithm is creatively proposed to estimate the component of the original signal.This method fully considers the morphological characteristics of the slices on the time-frequency support domain of the signal,and after multichannel accumulation,the components have greater energy on the slices and the peaks of the spectral lines are sharper.Therefore,the multichannel time-frequency slice accumulation algorithm can more accurately estimate the number of components of the original signal,making the MMD method a decomposition method with full adaptability.(4)An envelope order spectroscopy method based on the MMD method is proposed,and the MMD method is applied to the vibration signal of the planetary gearbox under variable speed working conditions and the vibration signal of the fixed-axis gearbox.Experimental data analysis shows that the method accurately estimates the number of original signal components,and accurately extracts the signal components and strong time-varying components of the adjacent distribution of the low frequency domain in the complex time-frequency support domain,and effectively extracts the signal fault characteristics of the variable working gear.Compared with other decomposition methods,the superiority of this method is shown.