This invention relates to the field of intelligent bearing monitoring technology, and discloses a method, device, equipment, and storage medium for detecting
bearing wear status. The method provided by this invention separates and extracts long-term trend components and short-term fluctuation components over a time scale, quantifying the remaining
lubricant thickness and interface stress fluctuations respectively, thus decoupling wear accumulation from transient friction behavior. Further
statistical analysis of short-term variance and pulse count rate reflects the uniformity of
stress distribution and the frequency of direct
metal-to-
metal contact. Normalized variance is used to eliminate the influence of operating conditions such as load and speed on
luminescence intensity, avoiding misjudgments caused by changes in operating conditions. Finally, by combining the above parameters and using a wear stage discrimination model, the wear stage discrimination result of the target bearing at the current moment can be accurately identified. This not only achieves non-invasive online monitoring of
bearing wear status in sealed, confined spaces, but also accurately identifies the
bearing wear status using changes in
light intensity in the original
triboluminescence signal of the target bearing.