TY - GEN
T1 - ANN-based Prediction of Nusselt Number and Stored Energy in PCM Heat Exchanger for Solar Heat Storage
AU - Nguyen, Thi Nhan
AU - Le, Van Cong
AU - Noh, Se Hyeon
AU - Park, Chan Woo
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - In this study, heat transfer during the melting process of a phase change material (PCM) is investigated. The PCM used in this study is paraffin wax. To predict Nusselt number (Nu), and stored energy (E), two artificial neural network (ANN) models were developed. Datasets collected from experiment are used for training, validating and testing the ANN model in the ratio of 80%, 10% and 10%, respectively. For predicting of Nusselt number, Rayleigh, Fourier and Stefan numbers are set as input parameters and the output is Nusselt number. For predicting stored energy, temperature, flow rate, and time is the input variable while stored energy is the output temperature. The accuracy of developed model is evaluated by Mean square error (MSE), R-square (R2) and Mean Absolute Percentage Error (MAPE). The optimal structure of ANN based on minimum MSE, high R2 and low MAPE. The accuracy of the two models were proved by 0.0026 MSE, 0.851 R2 and 0.128 MAPE for predicting Nu and 0.0003 MSE, 0.993 R2, 0.033 MAPE for predicting stored energy. Two ANN models have high accuracy in predicting the heat transfer characteristics and the energy stored during PCM melting.
AB - In this study, heat transfer during the melting process of a phase change material (PCM) is investigated. The PCM used in this study is paraffin wax. To predict Nusselt number (Nu), and stored energy (E), two artificial neural network (ANN) models were developed. Datasets collected from experiment are used for training, validating and testing the ANN model in the ratio of 80%, 10% and 10%, respectively. For predicting of Nusselt number, Rayleigh, Fourier and Stefan numbers are set as input parameters and the output is Nusselt number. For predicting stored energy, temperature, flow rate, and time is the input variable while stored energy is the output temperature. The accuracy of developed model is evaluated by Mean square error (MSE), R-square (R2) and Mean Absolute Percentage Error (MAPE). The optimal structure of ANN based on minimum MSE, high R2 and low MAPE. The accuracy of the two models were proved by 0.0026 MSE, 0.851 R2 and 0.128 MAPE for predicting Nu and 0.0003 MSE, 0.993 R2, 0.033 MAPE for predicting stored energy. Two ANN models have high accuracy in predicting the heat transfer characteristics and the energy stored during PCM melting.
KW - artificial neural network
KW - Nusselt number
KW - phase change material (PCM)
KW - solar heat storage
UR - https://www.scopus.com/pages/publications/85141386376
U2 - 10.1109/BCD54882.2022.9900585
DO - 10.1109/BCD54882.2022.9900585
M3 - Conference paper
AN - SCOPUS:85141386376
T3 - Proceedings - 2022 IEEE/ACIS 7th International Conference on Big Data, Cloud Computing, and Data Science, BCD 2022
SP - 94
EP - 98
BT - Proceedings - 2022 IEEE/ACIS 7th International Conference on Big Data, Cloud Computing, and Data Science, BCD 2022
A2 - Trong, Van Hung
A2 - Park, Jongwoo
A2 - Thao, Vo Thi Thanh
A2 - Kim, Jongbae
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 7th IEEE/ACIS International Conference on Big Data, Cloud Computing, and Data Science, BCD 2022
Y2 - 4 August 2022 through 6 August 2022
ER -