Performance analysis and modeling of bio-hydrogen recovery from agro-industrial wastewater

Significant volumes of wastewater are routinely generated during agro-industry processing, amounting to millions tonnes annually. In line with the circular economy concept, there could be a possibility simultaneously treating and recovering bio-energy resources such as bio-hydrogen. This study aimed model effect different process parameters that influence treatment recovery from agro-industrial wastewaters. Three wastewaters dairy, chicken palm oil mills were investigated. Eight data-driven machine learning algorithms namely linear support vector (LSVM), quadratic (QSVM), cubic (CSVM), fine Gaussian (FGSVM), binary neural network (BNN), rotation regression (RQGPR), exponential (EQGPR) (EGPR) employed for modeling process. The datasets obtained three processes train test models. LSVM, QSVM, CSVM did not show an impressive performance indicated by coefficient determination (R2) < 0.7 prediction hydrogen produced using processes. models also characterized high errors. Superior was displayed FGSVM, BNN, RQGPR, EQGPR, EQGPR R 2 > 0.9, indication better predictability minimized errors low root mean square error (RMSE), (MSE), absolute (MAE).