Research On Fault Diagnosis Of Rotating Machinery For Gearbox Of Vehicle

The study of gearbox problem has a great effect on the collection,denoising,selection of eigenvectors and fault judgments under special conditions such as high intensity disturbances.Aiming at the end effect of empirical mode decomposition(EMD)in the process of decomposition,the problem can be dealt with by using the even extension method and the cosine window function method.So the first step on the first signal on the even extension processing,the second step due to even after the existence of even delay error exists,so add a cosine window to the signal,so that this error can be controlled at both ends,so that It will not let the error and then expand,reduce the endpoint effect on the decomposition of the correct rate of impact.Finally,through the experiment to effectively verify that this approach can effectively reduce the impact of endpoint effects,improve the accuracy of fault determination.(Hilbert-AR method)and the autoregressive model(AR)collocation model(Hilbert-AR method)are proposed.In this way,the EMD method is used to decompose the vibration signal into a number of intrinsic modal functions(IMF),and then all the IMF is subjected to the Hilbert transform method to obtain the instantaneous frequency(IF).An AR model is established to select its coefficients and residual The variance of the matrix singular value as the eigenvector,the final experimental results verify that the method of extraction characteristics of high precision,the effect is better.