The Prediction of Sinter Drums Strength Using Hybrid Machine Learning Algorithms.

Area of Science:

• Materials Science
• Metallurgical Engineering
• Data Science

Background:

• Sinter drum strength is a critical quality indicator in sintering processes.
• Accurate prediction of sinter drum strength is essential for process optimization and quality control.

Purpose of the Study:

• To establish a machine learning-based prediction model for sinter drum strength.
• To identify optimal data preprocessing and regression algorithms for accurate sinter drum strength prediction.

Main Methods:

• Utilized historical sintering data to build regression prediction models.
• Evaluated ten machine learning regression algorithms, including linear regression, support vector regression, and random forest regression.
• Compared various data preprocessing techniques, focusing on data standardization.

Main Results:

• Data standardization was identified as the most effective preprocessing method.
• Gradient boosting regression, random forest regression, and extra tree regression demonstrated superior prediction accuracy.
• Optimized algorithm parameters for enhanced sinter drum strength prediction.

Conclusions:

• Machine learning models, particularly with standardized data and ensemble regression methods, can accurately predict sinter drum strength.
• The developed model provides a valuable tool for optimizing sintering processes and ensuring product quality.