A Study On The Local Fault Diagnosis Of Planetary Gearbox Based On Translocation Resampling And Improved VMD

The planetary gearbox is one of the important parts in rotating machinery.Because the planetary gearbox can not only reduce the transmission speed ratio,but also can be proportional to the increase of torque,so in the rotating machinery such as reducer,car transmission has a wide range of applications.Because the planetary gearbox needs to meet the complex working requirements of the mechanical equipment,it is generally in an off-stationary operating condition,resulting in more complex vibration signal components of the planetary gearbox.Planetary gear box is affected by subjective and objective factors in the working process,and its speed is often unsteady,which increases the difficulty of extracting fault features from vibration signals.Therefore,it is necessary to study the fault diagnosis method of variable speed planetary gearbox.For the variable speed planetary gear box under the non-stationary condition,the change of the speed will make the time frequency structure in the signal extremely complex,and make some fault symptoms become fuzzy or even indistinguishable.At this time,the steady-state signal processing method will lead to misjudgment.At the same time,the complex working environment of the planetary gearbox brings strong interference to the vibration signal,resulting in low SNR of the vibration signal of the gearbox,which further increases the difficulty of fault feature extraction.Therefore,an appropriate signal denoising method should be selected to process the signal and improve the signal to noise ratio.It is helpful to extract fault features and make fault judgment.In this paper,a method of fault diagnosis of planetary gearbox based on rotating domain resampling and improved VMD technology is presented for the characteristics of gear tooth fault signal of planetary gearbox with variable speed.The rotating domain resampling technique eliminates the influence of speed variation on fault characteristic frequency.The improved adaptive VMD method is used to decompose the order signal and improve the signal-to-noise ratio.Then combining d-s theory with VMD method,the intrinsic modal components obtained by VMD decomposition are selected by quantization parameters,and the IMF obtained by VMD treatment is evaluated by multi-index fusion,which reduces the accidental error of single index in signal component evaluation and improves the reliability of signal reconstruction.The fault diagnosis of variable speed planetary gearbox is realized by systematic method.(1)In traditional computing order analysis of the shortage problem of the accuracy of angular domain re-sampling,proposed transfer domain re-sampling resampling step in order to calculate time analysis was improved,to improve the precision of the order analysis,to improve the precision of the calculation order analysis,then turn domain resampling the Angle domain signal is converted to class set speed order signal;To solve the problem that the traditional time-domain and frequency-domain analysis methods cannot be applied to the variable-speed signal.(2)Using variational mode decomposition(VMD)will be converted order time signal filtering processing,solve the problems of strong interference signals in variable speed signal,using the energy method in the key parameters of VMD decomposition after evaluation of the original signal decomposition of energy and the consistency of the original signal,thereby further determine whether the degree of decomposition of VMD is appropriate,to avoid a breakdown and owe decomposition.After confirming the key parameter of VMD decomposition,namely the number of VMD decomposition layers,the signal was reconstructed.(3)For the components of the signal generated by the VMD decomposition,the multi-index fusion was used to quantitatively evaluate each component.Kurtosis,correlation coefficient and permutation entropy were selected to evaluate the decomposition component of the signal,and the impact,original signal correlation and complexity of the decomposed IMF were evaluated respectively.For the three selected parameters,the d-s theory of information fusion theory is used to weighted fusion of the three parameters,and the selected ruler of component evaluation is confirmed.After the evaluation of each component,whether it is suitable for reconstruction is confirmed.After the quantified evaluation results are obtained,the components satisfying the ruler are reconstructed.Fault identification of reconstructed signals.