Development of a deep learning predictive model for estimating higher heating value in municipal solid waste management

Abstract

The management of municipal solid waste (MSW) has become a pressing issue in urban areas due to population growth and increasing waste generation. Waste-to-energy (WTE) conversion through thermo-chemical processes offers a promising solution, where the Higher Heating Value (HHV) of MSW plays a crucial role in process optimization. Traditional calorimetric methods for HHV determination are labor-intensive, costly, and environmentally harmful, prompting the need for automated, efficient predictive models. In this work, a novel deep learning-based framework called DLHHV-MSW is presented it estimates the HHV of MSW from its elemental composition, such as the amount of ash, carbon, hydrogen, nitrogen, oxygen, sulfur, and water. The framework utilizes a Deep Belief Network (DBN) optimized with Oppositional Cat Swarm Optimization (OCSO) to improve predictive accuracy. Experimental results demonstrate that DLHHV-MSW achieves superior performance, with a correlation coefficient (CC) of 0.996 and a mean squared error (MSE) of 2.342, outperforming traditional methods. This automated approach offers a scalable, cost-effective, and environmentally friendly solution for enhancing WTE operations and advancing sustainable MSW management.