Bearing fault diagnosis under variable speed conditions on adaptive time frequency extraction mode decomposition

Abstract

Fault diagnosis of rolling bearings is crucial for their operation in variable working conditions in mechanical systems. Time-frequency distribution (TFD) methods are widely used in variable-speed rolling bearing fault diagnosis due to their ability to capture the time-frequency characteristics of signals. However, they suffer from weak robustness, poor resolution, and lack of feature information concentration. To address these issues, we construct a method: adaptive time frequency extraction mode decomposition (ATFEMD) by utilizing the characteristic of instantaneous frequency (IF) energy concentration in TFD. The method first enhances the adaptive time-frequency distribution (ADTFD) via an iterative algorithm, applying random sample consensus (RANSAC) algorithm enables ridges extraction from the TFD. The harmonics of the extracted ridges are then verified by passing the proxy data method, followed by a noise test to determine if the harmonics are true. From the above steps, ATFEMD decomposing the signal into physically meaningful adaptive mode (AM). This article analyzes the simulated bearing vibration and test bench bearing vibration signals to demonstrate the performance of ATFEMD. Results indicated that the proposed method is characterized by strong robustness, and good feature extraction results compared to other methods.