Modeling the gasification process in producing raw gases and solids using machine learning techniques in combination with optimization algorithms

The research deals with the development of artificial intelligence models for the prediction of biomass gasification systems to achieve better efficiency and reliability. Conserve environmental health, different input data for the modeling of gasification output materials are used, such as gases and solids, and reduce reliance on fossil fuel supplies. It adopted the development of hybrid frameworks that coupled an Extra Tree Regressor (ET) model with two optimization algorithms, namely Equilibrium Slime Mold Algorithm (ESMA) and Manta Ray Foraging Optimization (MRFO). The dataset from 312 experiments found in the literature was employed for the simulation of the gasification process. The obtained result manifested that the ETES hybrid model had an excellent fitness accuracy with R2 = 0.99 and RMSE = 0.0694 for gas yield (GY) and R2 = 0.986 and RMSE = 6.943 for char yield (CY) during the training phase, respectively. These results confirm that the hybrid models can remarkably improve the prediction of both syngas and char from gasification processes. The implications of this research are overwhelming, as it would render a reliable tool for operational management in gasification processes, contributing to the production of clean fuel and supporting sustainable development. This might allow, after further refinement, additional research, and experimentation on these models to increase their precision and applicability for a wide range of gasification scenarios.