Prediction of thermal conversion characteristics for co‐combustion of waste tire–lignite coal using machine learning algorithms

Abstract Co‐combustion of coal with various waste resources is an effective energy recovery strategy that integrates waste‐derived fuels while reducing dependence on fossil fuels. In this study, machine learning algorithms were used to predict thermogravimetric data for the co‐combustion process of waste tire (WT) and lignite coal (LC) blends to improve the understanding of the thermal conversion characteristics. The study analyzed the combustion behaviour of WT, LC, and their mixtures at four different heating rates (10, 20, 30, and 40°C/min) and various mixing ratios (100:0, 20:80, 40:60, 50:50, 60:40, 80:20, and 0:100) using thermogravimetry–derivative thermogravimetry/differential scanning calorimetry (TG‐DTG/DSC) techniques. To improve the prediction accuracy, eight machine learning algorithms—adaptive boosting regression, decision tree regression, k‐nearest neighbour regression, linear regression, multi‐layer perceptron, random forest regression, support vector machine regression, and XGBoost—were applied to model the co‐combustion process. The results showed a strong correlation between experimental data and machine learning predictions, confirming the effectiveness of these models. By enabling accurate real‐time prediction of thermal conversion characteristics, this study reduces the reliance on labour‐intensive thermogravimetric analysis (TGA) and facilitates cost‐effective, adaptive, and scalable optimization of combustion processes for industrial applications.