Scholarlink

Bearings play a crucial role in any rotating machine. Remaining Useful Life (RUL) of the bearing is required to be accurately predicted for the effective maintenance in industries to avoid abrupt shutdowns. To expedite the development of an automated RUL prediction model, development of a Health Indicator (HI) that effectively and precisely describes the deterioration process of bearing is important. This work utilizes the monotonicity score from features to construct the HI of each bearing to predict accurate RUL. To find out the bearing’s First Predicting Time (FPT), a 3-sigma criterion determined by the Root Mean Square (RMS) of the vibration signal is used. The RUL of bearing is predicted using Gaussian Process Regression (GPR). The proposed methodology is tested using an open-source XJTU–SY dataset. The finding shows that the monotonicity score helps in constructing effective HI which improves the RUL prediction accuracy.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Monotonicity Score-Based Remaining Useful Life Prediction of Bearing Using Gaussian Process Regression

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Afroz Ahmed Saad,
  • Anurag Choudhary,
  • Shahab Fatima,
  • Bijaya Ketan Panigrahi

摘要

Bearings play a crucial role in any rotating machine. Remaining Useful Life (RUL) of the bearing is required to be accurately predicted for the effective maintenance in industries to avoid abrupt shutdowns. To expedite the development of an automated RUL prediction model, development of a Health Indicator (HI) that effectively and precisely describes the deterioration process of bearing is important. This work utilizes the monotonicity score from features to construct the HI of each bearing to predict accurate RUL. To find out the bearing’s First Predicting Time (FPT), a 3-sigma criterion determined by the Root Mean Square (RMS) of the vibration signal is used. The RUL of bearing is predicted using Gaussian Process Regression (GPR). The proposed methodology is tested using an open-source XJTU–SY dataset. The finding shows that the monotonicity score helps in constructing effective HI which improves the RUL prediction accuracy.