Enhancement of the AR-MED Deconvolution Process using Ensemble Empirical Mode Decomposition in Bearing Fault Detection

Due to its ability to enhance the fault impulses,autoregressive minimum entropy deconvolution (AR-MED) is a useful method in bearing fault diagnosis. However, when the signal includes a high level of noise, it seriously affects the deconvolution performance of AR-MED. Therefore, early defect information is hardly extractable from the vibration signals of rolling bearings. This paper aims to study the effectiveness of integrating ensemble empirical mode decomposition (EEMD) with AR-MED to improve the deconvolution procedure. Firstly, using EEMD, it is possible to divide the defect signal into IMFs. Then the relevant modes that contain defect information are identified via the IMFs envelope spectrum. After that, AR-MED is applied to the reconstructed signal. Finally, the results obtained are compared with those obtained by deconvolving directly the original vibration signal using AR-MED. The experimental results show that incorporating EEMD improves the deconvolution process of AR-MED for bearing fault diagnosis.

• bearing fault diagnosis
• ensemble empirical mode decomposition
• rolling bearings
• vibration signals
• noise reduction
• intrinsic mode functions
• envelope spectrum

• Machine Fault Diagnosis Techniques
• Gear and Bearing Dynamics Analysis
• Fault Detection and Control Systems
• bearing fault diagnosis
• ensemble empirical mode decomposition
• rolling bearings
• vibration signals
• noise reduction
• intrinsic mode functions
• envelope spectrum
• Control and Systems Engineering
• Mechanical Engineering
• Engineering