Indicateurs pour le pronostic de la durée de vie utile restante (RUL) des roulements

Abstract

Bearings are one of the most used and easily damaged components in rotating machinery. Therefore, the health status of these sensitive components directly affects the performance of all equipment. In order to avoid accidents, failures and economic losses, it is necessary to accurately diagnose bearing failures over time and take corresponding maintenance actions based on the diagnostic results. The health prediction and management (PHM) method provides information about the health of various components of a mechanical system (such as bearings) and estimates their life (remaining life, RUL) before failure. In data-driven forecasting methods, indicator information plays an important role in the reliability of the forecast. Although this work interested in developing effective methods to select the best degradation indicators