OPTIMIZATION OF ENERGY CONSUMPTION IN CNC MARBLE PROCESSING: STATISTICAL AND MACHINE LEARNING APPROACHES

This study aims to evaluate the specific energy consumption during marble processing on CNC machines both by traditional statistical methods and machine learning models. It presents an analytical framework that examines the effects of process parameters to improve energy efficiency in CNC machining processes. In the experimental part, a data set of 5400 obrervations was obtained considering different machining types, depths of cut and feed rates. Analysis of Variance (ANOVA) and regression models confirmed the decisive role of material removal rate (MRR) on specific energy consumption. The study comprehensively analyzed the performance of four different machine learning models (Gradient Boosting, Random Forest, XGBoost, LightGBM) to predict the specific energy consumption during marble processing on CNC machines. The findings show that specific energy consumption is an important parameter for energy efficiency and cost reduction. The accuracy of the models was evaluated with metrics such as R2, RMSE and MAE, and as a result, it was found that Gradient Boosting and XGBoost models outperformed the others in the Spiral machining type. These findings provide a solid basis for developing strategies to improve energy efficiency in marble processing on CNC machines. The study provides important information that can help make strategic decisions to save energy and improve environmental sustainability. Providing valuable guidance for future research, this study demonstrates the potential use of machine learning models to improve energy efficiency in the natural stone industry.