Development of a kiln petcoke mill predictive model based on a multi-regression XGBoost algorithm

• Published in: International journal of advanced manufacturing technology Vol. 130; no. 7-8; pp. 3373 - 3386
• Main Authors: Benchekroun, Mohammed Toum, Zaki, Smail, Aboussaleh, Mohamed, Belrhiti, Hajar, Diassana, Fatoumata
• Format: Journal Article
• Language: English
• Published: London Springer London 01.02.2024; Springer Nature B.V
• Subjects: Advanced manufacturing technologies; Algorithms; Artificial intelligence; CAE) and Design; Cement industry; Cement plants; Coal; Computer-Aided Engineering (CAD; Costs; Data analysis; Datasets; Electricity; Energy consumption; Energy efficiency; Engineering; Industrial and Production Engineering; Investigations; Literature reviews; Machine learning; Manufacturing; Mechanical Engineering; Media Management; Optimization; Original Article; Petroleum coke; Prediction models; Regression models; score; Multi-regression; RMSE; XGBoost; Petcoke mill
• ISSN: 0268-3768; 1433-3015
• DOI: 10.1007/s00170-023-12689-z

This paper presents an investigation into the optimization of petroleum coke mill or petcoke mill processes, to improve efficiency and reduce waste in the heavy industry within the cement plant where our study is conducted. Our mission was to create a robust algorithm that could properly anticipate the mill’s performance and improve its operations. To accomplish this, we started by performing a comprehensive data analysis. Next, we built numerous regression models, and then assessed the effectiveness of each model using four crucial metrics. The suggested model is a multi-regression XGBoost (eXtreme gradient boosting) model, performing with a 90% score. Finally, the model will then be used to build an algorithm that can optimize the input values to accomplish the intended results.