TY - GEN
T1 - Neural network based estimation of electricity generated during a waste-to-energy process
T2 - 6th International Conference on Soft Computing and Machine Intelligence, ISCMI 2019
AU - Ighravwe, Desmond Eseoghene
AU - Mashao, Daniel
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/11
Y1 - 2019/11
N2 - Many emerging economies are embarking on the production of electricity from food wastes. And this has rekindled the interest of waste-to-energy engineers in these economies. They are confident that machining learning algorithms will help them to reduce the computation cost for this process. Here, an artificial neural network (ANN) model is used to estimate the amount of electricity generated during a waste-to-energy process. The selected model is a single hidden layer model with five inputs including methane gas, compression efficiency, boiler efficiency and more-the model's output is electricity generated. This study evaluated ten ANN architectures for the prediction purpose; data from nine cities in Nigeria were used to achieve this purpose. The results obtained show that a 5-4-1 ANN architecture performs better than the other architectures during their training and testing phases. This model's training and testing mean square error is 6.96 x 10-5 and testing 3.62 x 10-5, respectively. Based on the ANN performance, it was concluded that it can be used to monitor a waste-to-energy process.
AB - Many emerging economies are embarking on the production of electricity from food wastes. And this has rekindled the interest of waste-to-energy engineers in these economies. They are confident that machining learning algorithms will help them to reduce the computation cost for this process. Here, an artificial neural network (ANN) model is used to estimate the amount of electricity generated during a waste-to-energy process. The selected model is a single hidden layer model with five inputs including methane gas, compression efficiency, boiler efficiency and more-the model's output is electricity generated. This study evaluated ten ANN architectures for the prediction purpose; data from nine cities in Nigeria were used to achieve this purpose. The results obtained show that a 5-4-1 ANN architecture performs better than the other architectures during their training and testing phases. This model's training and testing mean square error is 6.96 x 10-5 and testing 3.62 x 10-5, respectively. Based on the ANN performance, it was concluded that it can be used to monitor a waste-to-energy process.
KW - Artificial neural network
KW - Developing countries
KW - Electricity generation
KW - Waste-to-energy
UR - http://www.scopus.com/inward/record.url?scp=85081581030&partnerID=8YFLogxK
U2 - 10.1109/ISCMI47871.2019.9004303
DO - 10.1109/ISCMI47871.2019.9004303
M3 - Conference contribution
AN - SCOPUS:85081581030
T3 - 2019 6th International Conference on Soft Computing and Machine Intelligence, ISCMI 2019
SP - 56
EP - 60
BT - 2019 6th International Conference on Soft Computing and Machine Intelligence, ISCMI 2019
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 19 November 2019 through 20 November 2019
ER -
