Adaptive neuro-fuzzy inference system (ANFIS) approach for the irreversibility analysis of a domestic refrigerator system using LPG/TiO2 nanolubricant
Loading...
Date
2020-05-28
Journal Title
Journal ISSN
Volume Title
Publisher
Energy Reports
Abstract
This work presents an adaptive neuro-fuzzy inference system (ANFIS) artificial intelligence methodology
of predicting the 2nd law efficiency and total irreversibility of a refrigeration system running
on LPG/TiO2–nano-refrigerants. For this purpose, substractive clustering and grid partition approaches
were utilized to train the ANFIS models required in estimating the 2nd law efficiency and total
irreversibility using some experimental data. Furthermore, predictions of ANFIS models with subtractive
clustering approach was found to be more accurate than ANFIS models predictions with grid
partition approach. The predictions of ANFIS models with subtractive clustering approach were also
compared with experimental results that were not included in the model training and predictions
of already existing ANN models of authors previous publication. The comparison of variance, root
mean square error (RMSE), mean absolute percentage error (MAPE) were 0.996–0.999, 0.0296–0.1726
W and 0.108–0.176 % marginal variability values. These results indicate that the ANFIS model with
subtractive clustering approach having cluster radii 0.7 and 0.5 can predict the 2nd law efficiency and
total irreversibility respectively, with higher accuracy than authors’ previous publication ANN models.
Description
Staff Publication
Keywords
LPG, ANN, TiO2nanoparticle, Total irreversibility, ANFIS, 2nd law efficiency