Modeling of Bio-Oil Production by Pyrolysis of Woody Biomass: Artificial Neural Network Approach

Authors : Günay ÖZBAY, Erkan Sami KÖKTEN

Pages : 1255-1264

Doi:10.2339/politeknik.659136

View : 35 | Download : 10

Publication Date : 2020-12-01

Article Type : Research Paper

Abstract :This study is dedicated to developing a reliable artificial neural network insert ignore into journalissuearticles values(ANN); model to model the pyrolysis liquid product insert ignore into journalissuearticles values(bio-oil);. Some related parameters with the bio-oil yield such as the pyrolysis temperature, duration, catalyst type, catalyst ratio, particle size, proximate, and ultimate analysis of the biomass were tested. Due to the different characteristics of different biomass types and pyrolysis methods, only slow and intermediate pyrolysis data from woody biomass were used in modeling. The correlation coefficients insert ignore into journalissuearticles values(R); were 0.992, 0.933, and 0.951 for training, validation, and testing, respectively. In order to evaluate the predictability of the ANN model, the predicted results were compared with the experimental results that were not introduced before. The simulated data were in good agreement with the experimental results indicating the reliability of the developed model. The relative impact results revealed that the most important parameter that affects the bio-oil yield was catalyst type insert ignore into journalissuearticles values(11.4%);.  
Keywords : catalyst, modeling, pyrolysis, Artificial neural network, bio oil

ORIGINAL ARTICLE URL