A novel time series model-based indicator for bearing degradation monitoring

As a key component in complex mechatronic systems, rolling element bearing (REB) degradation is often inevitable and can cause catastrophic failures. Hence, it is critical to monitor bearing degradation through online techniques. However, conventional condition monitoring health indicators (HIs) are ineffective in detecting faults at an early stage and are insufficient in quantifying defects. This paper proposes HIs based on the generalised autoregressive conditional heteroskedasticity (GARCH) time series model to characterise the evolution of the dynamic response of an REB system, specifically the cyclostationarity of the repetitive transients. The validity of the proposed indicators is assessed using a publicly available dataset and acoustic emission (AE) signals acquired from an accelerated bearing degradation test. The result shows that the GARCH-based indicators can determine the incipient failure earlier than traditional statistical HIs and have the ability to quantify failure to some extent. In addition, the proposed indicators have the potential to be employed in studies pertaining to the prediction of remaining useful life.