Abstract
In response to issues such as the lack of capability for timely early warning and the difficulty in monitoring the status of rolling bearings, a condition-monitoring method for rolling bearings based on the Honey Badger Algorithm (HBA) for optimizing dynamic asynchronous periods is proposed. This method is founded on the peak factor and involves comparing peak factors at different periods to construct a dynamic asynchronous peak-factor-ratio-monitoring index, which is then optimized using the HBA. Simulated experiments were carried out using the XJTU-SY dataset. The results indicate that, compared to the early warning times defined by international standards, the warning times provided using this method are consistently over 33 min in advance within the test dataset. Additionally, an envelope spectrum analysis of the warning data confirms the existence of early faults. This demonstrates that the monitoring indicator developed in this paper is capable of delivering earlier and more accurate early fault warnings and condition monitoring for rolling bearings.