Degradation Modeling and RUL Prediction of Hot Rolling Work Rolls Based on Improved Wiener Process.

Area of Science:

• Materials Science and Engineering
• Mechanical Engineering
• Industrial Engineering

Background:

• Hot rolling work rolls face severe degradation due to high temperatures, stress, and wear.
• Roll surface degradation critically impacts final product quality and operational efficiency.
• Accurate remaining useful life (RUL) prediction is vital for optimizing maintenance.

Purpose of the Study:

• To develop an improved degradation model for predicting the RUL of hot rolling work rolls.
• To enhance the accuracy and reliability of RUL estimation for industrial applications.
• To optimize maintenance strategies and ensure operational efficiency.

Main Methods:

• An improved degradation model based on the Wiener process.
• Integration of pulsed eddy current testing with VMD-Hilbert feature extraction.
• Incorporation of a Gaussian kernel into the Wiener process and use of a Bayesian framework for parameter estimation.

Main Results:

• The proposed model demonstrated significant reductions in RMSE (approx. 85.47% and 41.20%) and MAE (approx. 85.66% and 42.61%) compared to existing methods.
• Achieved substantial improvements in the coefficient of determination (CD) by 121% and 19.76%.
• Most RUL predictions fell within a 20% confidence interval, indicating high accuracy and robustness.

Conclusions:

• The improved Wiener process model offers superior predictive accuracy and robustness for RUL estimation of hot rolling work rolls.
• The model provides a reliable tool for real-time RUL prediction, optimizing maintenance and operational efficiency.
• This approach addresses the critical need for accurate roll degradation assessment in industrial settings.