Forecasting gasification sustainability through enhanced K-nearest neighbour models for hydrogen and nitrogen amount

Abstract This study pioneers a method for forecasting hydrogen and nitrogen quantities in gasification processes, which is crucial for transforming carbon-based materials into valuable gases with minimal environmental impact. It addresses the pressing need for precise and economical solutions in gasification by streamlining estimation procedures across various operational conditions. Anchored in historical data, the K-nearest neighbour’s (KNN) model forms the core of this method, adeptly capturing intricate relationships between input variables and gas production results. What sets this research apart is the integration of advanced optimization techniques, Beluga whale optimization (BWO), and Northern Goshawk Optimization (NGO), further refining the accuracy of the KNN model’s predictions. This work signifies a substantial stride in optimizing gasification processes, offering a pathway towards more efficient and sustainable conversion of carbon-based materials, showcasing the potential of data-driven sustainability in this domain, and ultimately diminishing the environmental impact of these operations. Results highlight the outstanding predictive performance of the KNNG model, achieving an impressive R 2 value of 0.995 and 0.994 during training. Notably, both the KNNG and KNBW models outperformed the basic KNN model in forecasting gasification outputs, underscoring the reliability and effectiveness of these optimized models in predicting these processes.