BEARING HEALTH ASSESSMENT AND FAULT DIAGNOSIS USING THE METHODS OF SELF-ORGANIZING MAP

摘要

This paper presents a method for bearing health assessment and fault diagnosis, which is comprised of the following two steps: (1) Fast Fourier Transform (FFT) technique is applied to vibration signals to extract feature vectors, each of which describing a state of bearing health; (2) the well-known neural network technique Self-Organizing Map (SOM) method is applied to the extracted feature vectors to draw conclusions concerning the health condition of a bearing. The presented method is illustrated and validated through an industrial case study.