Heating value prediction model of Acacia mangium Willd using near infrared spectroscopy

Abstract

This study reported the prediction model of heating value for the Acacia mangium Willd which is the promoted as energy plant for the farmer to grow in the farm. Its heating value is approximately of 19 kJ/kg which provides high potential using for an alternative energy as biomass. The near infrared spectroscopy technique (NIR) is used to create model to predict the heating value of this biomass in order to reduce the investigated time. The results of model on the validation set were the coefficient of determination (R 2 ) of 71%, RMSEP of 246 J/g, RPD 1.87 and bias of 33.7 J/g. These results showed the potential and possibility to apply the NIR prediction technique. The robust model was suggested to carried out for more accuracy and may be applied in the online system analysis.