BIOETHANOL PRODUCTION FROM ABIES ALBA WOOD USING ADAPTIVE NEURAL FUZZY INTERFERENCE SYSTEM MATHEMATICAL MODELING

This study investigates bioethanol production from Abies alba wood. The wood was first autohydrolysed, then delignified and the remaining cellulose was used as substrate for simultaneous saccharification and fermentation processes. The influence of temperature (180, 190 and 200 C) and pretreatment time (5, 10 and 15 min) on the fermentation medium products was studied. The maximal bioethanol content (52.0 g L -1 ) was obtained at a pretreatment temperature of 190 C and pretreatment time of 10 min. The enzymatic hydrolysis and fermentation temperature was 38 C for 72 h. The untreated, autohydrolysed and delignified wood was characterized by reflected light microscopy for morphological structure identification. The adaptive network-based fuzzy interference model (ANFIS) and the Gaussian membership function were used to reproduce the experimental results obtained for complete characterization of the wood fermentation broth. The proposed model uses two input variables (temperature and reaction time) and one output parameter based on two intelligent methods: back-propagation and a hybrid method. The hybrid intelligent method has good accuracy (99.2-100.0%) and correlation coefficient (0.998-1). The fermentation broth contains a mixture of bioethanol and secondary by-products, including acids, alcohols, aldehydes, ketones and esters. A maximum of 5.2 g bioethanol can be obtained from 100 g of woody biomass after autohydrolysisdelignification-SSF process.