Cubic Spline Extreme Point Interpolation And Wavelet Filtering Are Used In The Fault Diagnosis Of Gear Partial Broken Teeth

Gear has become very important in machinery with a kind of widely used transmission equipment,metallurgy and other national economy industries.When gear failure occurs,it will affect normal production.It is very important to know the running quality of gear.Therefore,it is of great significance to study the fault diagnosis of gear vibration signal.The purpose of analyzing gear vibration signal is to extract fault features and make fault judgment.In the working state of gears,with a large amount of noise,the resulting vibration signal has noise-bearing and non-stationary.Therefore,when the gear vibration signal is analyzed,it is necessary to carry out relevant noise reduction filtering and adaptive stationary processing.In this paper,aiming at the noise,non-stationarity,noise reduction and adaptive stationarity of gear vibration signal,the extreme Point Packet Mode Decom is decomposed from wavelet and extreme point fitting based on cubic spline extremum.Positioning EPPMD was introduced into the fault diagnosis of gear partial broken teeth,and the validity of the method was verified by relevant experiments.The main contents of this paper are as follows:(1)The wavelet analysis is applied to the fault diagnosis of gear partial broken teeth.The window of wavelet analysis is changed,and the localization of the gear is also changed.1)in view of the large amount of noise in the vibration signal of gear,the wavelet analysis is applied to the fault diagnosis of gear local broken teeth.It has high time resolution at high frequency and high frequency resolution at low frequency,but noise often shows high frequency part,so it is suitable for noise analysis because wavelet threshold can be used to reduce noise.(2)Aiming at the non-stationarity of gear vibration signal,which affects the analysis of gear vibration signal and the extraction of fault feature,the adaptive method will be used to extract the signal at various scales.The method of EPPMD is applied to the diagnosis of gear partial broken tooth fault.EPPMD has the same adaptability as EMD,and can decompose the signal into IMF components with different scales.(3)Aiming at the proposed EPPMD method,the description of the method is introduced.When the signal is analyzed and processed,the extreme point of the signal is extracted first,and the extreme point of the extracted signal is grouped.Then the extreme points of each group are fitted with cubic splines,and the mean curves of all the group curves are obtained.Finally,the original signal subtracts the mean curve.Decomposing the original signal into IMF component Curves with different time Scales.The method of EPPMD generalizes the EMD method.The extreme points are divided into two groups in EPPMD,and then the extreme points of each group are fitted with cubic spline to become the top of the EMD.Lower envelope,so EPPMD contains EMD and is validated in test analysis.(4)Aiming at the fact that the EPPMD is essentially a filter,the spectrum characteristics of the IMF component of random noise processed by EPPMD are analyzed.It is found that the spectrum range of the former IMF component is equal to that of the latter,and that the spectrum range of the former IMF component is equal to that of the latter.It is verified that the nature of EMD is a binary filter with a spectral metric ratio of 2.(5)Aiming at the noise and non-stationarity of gear local tooth broken fault signal,a method of combining wavelet filter with the decomposition of extreme point grouping mode based on cubic spline extreme point fitting is proposed to analyze the gear local tooth broken fault test.The validity of the method is verified.