Low-Complexity Vibration-Based Condition Monitoring for Rolling Bearings: A Novel Instantaneous Amplitude–Frequency Approach
In vibration-based condition monitoring (VBCM) of rotating machinery, bearing defects typically induce amplitude and frequency modulations, making joint instantaneous amplitude–frequency analysis effective for distinguishing healthy from abnormal vibration patterns. Recent advances in sen...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Open Journal of Instrumentation and Measurement |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11082271/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In vibration-based condition monitoring (VBCM) of rotating machinery, bearing defects typically induce amplitude and frequency modulations, making joint instantaneous amplitude–frequency analysis effective for distinguishing healthy from abnormal vibration patterns. Recent advances in sensor technology have enabled the deployment of smart transducers for VBCM, allowing vibration signals to be processed directly at the measurement point. However, the limited processing capabilities of these transducers necessitate efficient methods for signal analysis and feature extraction. To address this need, we propose a low-complexity, parameterless method for VBCM of rolling bearings based on joint instantaneous amplitude–frequency analysis. The method leverages the instantaneous amplitude (envelope) and instantaneous frequency of the vibration signal to construct two novel representations: 1) instantaneous amplitude–frequency mapping (IAFM) and 2) instantaneous amplitude–frequency correlation (IAFC). These representations preserve temporal information and capture condition-specific energy–frequency variations. Consequently, five new defect-sensitive features are extracted to characterize these variations. Experimental results demonstrate excellent performance in detecting and diagnosing various fault types across different motor speeds, confirming the method’s effectiveness in distinguishing healthy and faulty states. Additionally, its moderate computational complexity makes the method well suited for real-time monitoring using smart transducers. |
|---|---|
| ISSN: | 2768-7236 |