By quantifying the regularity vibration signals measured on rolling bearings, Approximate Entropy (ApEn) provides an effective measure for characterizing the structural degradation of bearings, and the severity of the defect. This paper investigates the relationship between ApEn and different failure modes of bearings. It is shown that. ApEn values decrease with the degradation of bearing defects. After introducing the theoretical background, experimental analysis are presented to quantify variation of ApEn values as a measure for defect mode and severity. A life cycle experiment is introduced to evaluate a defect growth precondition model based on regression analysis and Genetic Algorithm. Results show that ApEn is effective for bearing defect diagnosis and remaining service life prognosis.

This content is only available via PDF.
You do not currently have access to this content.