Engine fault feature extraction based on order tracking and VMD in transient conditions

Abstract

A method based on order tracking and Variational Mode Decomposition was proposed to solve the problem of non-stationarity and background noise in the vibration signal of engine. For the vibration signal of diesel engine crankshaft bearing fault and Connecting rod bearing of gasoline engine, the non-stationary signal in the time domain was converted into a pseudo stationary signal on the angular domain based on order tracking. Then the reconstructed signal was decomposed by several components by VMD, and the component that contains fault information was selected. The three-dimensional spectral array of order, speed and power spectrum was calculated, and the fault feature was extracted. The effectiveness of the proposed method is verified by simulation and experiment.