Integration of the grey relational analysis with machine learning for sucrose anaerobic hydrogen production prediction

Abstract

Anaerobic fermentation for hydrogen production is influenced by various environmental factors that can limit microbial activity. However, machine learning (ML) shows significant potential in explaining the complexity of biological processes. This study focuses on the application of ML in predicting hydrogen production rates (HPR) during anaerobic fermentation of biomass energy. Grey relation analysis was employed to investigate the correlation between operational parameters and HPR. Five ML algorithms, namely decision tree (DT), random forest (RF), extreme gradient boosting (XGBoost), and K-nearest neighbor (KNN), were utilized in conjunction with operating conditions and digestate characteristics as features, with HPR as the target variable. The evaluation of the models was conducted using mean squared error (MSE) and R squared (R2). Notably, butyrate, oxidation-reduction potential (ORP), and volatile suspended solids (VSS) were identified as key factors influencing hydrogen production during sucrose anaerobic fermentation. Among the ML algorithms tested, XGBoost achieved the highest accuracy, with an R2 value of 0.91 and an MSE of 0.0052.