TY - GEN
T1 - A GA-ANFIS Model for the Prediction of Biomass Elemental Properties
AU - Olatunji, Obafemi O.
AU - Akinlabi, Stephen
AU - Madushele, Nkosinathi
AU - Adedeji, Paul A.
N1 - Publisher Copyright:
© 2021, Springer Nature Singapore Pte Ltd.
PY - 2021
Y1 - 2021
N2 - The elemental composition of biomass is a significant property, which determines the energy content of biomass feedstock. This article develops a prediction model based on a hybrid adaptive neuro-fuzzy inference system (ANFIS) optimized with genetic algorithm (GA). The model inputs were the proximate constituents of biomass which are ash, fixed carbon, and volatile matter. These were used to predict the hydrogen (H), oxygen (O) and carbon (C) content of biomass fuels. The proposed algorithm was evaluated based on some known performance metrics. The root mean squared error (RMSE), mean absolute deviation (MAD), mean absolute percentage error (MAPE), coefficient of correlation (CC), mean absolute error (MAE) are 3.673, 2.4609, 5.1757, 0.9464, 0.309 at computation time (CT) of 33.65 secs for carbon (C); 0.6293, 0.4168, 8.3011, 0.75581, 0.0716 at CT of 40.21 secs for hydrogen (H); 4.4538, 3.1042, 13.3983,0.9167, 0.9899 at CT of 33.57 secs for oxygen (O), respectively. Regression analysis was also carried out to determine the level of dependence among the correlated variables. The model performance shows that GA-ANFIS can be applied in the computation of the elemental composition of biomass for strategic decision-making.
AB - The elemental composition of biomass is a significant property, which determines the energy content of biomass feedstock. This article develops a prediction model based on a hybrid adaptive neuro-fuzzy inference system (ANFIS) optimized with genetic algorithm (GA). The model inputs were the proximate constituents of biomass which are ash, fixed carbon, and volatile matter. These were used to predict the hydrogen (H), oxygen (O) and carbon (C) content of biomass fuels. The proposed algorithm was evaluated based on some known performance metrics. The root mean squared error (RMSE), mean absolute deviation (MAD), mean absolute percentage error (MAPE), coefficient of correlation (CC), mean absolute error (MAE) are 3.673, 2.4609, 5.1757, 0.9464, 0.309 at computation time (CT) of 33.65 secs for carbon (C); 0.6293, 0.4168, 8.3011, 0.75581, 0.0716 at CT of 40.21 secs for hydrogen (H); 4.4538, 3.1042, 13.3983,0.9167, 0.9899 at CT of 33.57 secs for oxygen (O), respectively. Regression analysis was also carried out to determine the level of dependence among the correlated variables. The model performance shows that GA-ANFIS can be applied in the computation of the elemental composition of biomass for strategic decision-making.
KW - Biomass feedstock
KW - Efficient utilization
KW - Elemental composition
KW - GA-ANFIS
UR - http://www.scopus.com/inward/record.url?scp=85090536252&partnerID=8YFLogxK
U2 - 10.1007/978-981-15-4745-4_95
DO - 10.1007/978-981-15-4745-4_95
M3 - Conference contribution
AN - SCOPUS:85090536252
SN - 9789811547447
T3 - Lecture Notes in Mechanical Engineering
SP - 1099
EP - 1114
BT - Trends in Manufacturing and Engineering Management - Select Proceedings of ICMechD 2019
A2 - Vijayan, S.
A2 - Subramanian, Nachiappan
A2 - Sankaranarayanasamy, K.
PB - Springer Science and Business Media Deutschland GmbH
T2 - 2nd International Conference on Mechanical Engineering Design, ICMechD 2019
Y2 - 25 April 2019 through 26 April 2019
ER -
