In-Sight Evaluation of Municipal Solid Waste Properties to Use as Green Fuel in a Waste to Energy Plant

This study aims to predict the higher heating value (HHV) of municipal solid waste (MSW) from the physical composition of MSW by using supervised machine learning. A total of 129 datasets of the physical composition of MSW collected throughout Thailand were used as a case study. Several mathematical models in MATLAB were selected for HHV prediction, including linear regression model (LRM), decision trees model (DTM), support vector machines model (SVM), efficient linear model (ELM), Gaussian process regression model (GPRM), kernel model (KM), and ensemble model (EM). Statistical analysis in this study considered a coefficient of determination (R2) as the "goodness of fit," and the results demonstrated that the EM could achieve the best R2-value of 0.57357. Thus, the EM was identified as the most appropriate model for determining the HHV of MSW based on the specified input parameters. Ultimately, the results of this study highlight that the EM and the comparative analysis of different models have the potential to predict the HHV from the physical composition of MSW, which offers valuable opportunities in the design and development of waste-to-energy (WtE) plants.