Abstract
The calculation of the performance of absorption heat transformers (AHTs) depends on multiple variables. In this work, artificial neural network (ANN) models with new configurations were developed to simultaneously estimate the coefficient of performance (COP) and Carnot coefficient of performance (COPCarnot) of an AHT prototype. The variables used to train the models were: the inlet and outlet temperatures corresponding to the main components of the AHT. The output parameters to simulate were the COP and COPcarnot, which are important values to determine the performance and real efficiency based on the Carnot cycle, respectively. To find the appropriate model, it was necessary to explore learning algorithms, activation functions, and multilayers. The results show a good estimation of the output parameters through three configurations of the ANN model. However, based on the number of coefficients obtained during learning and the simultaneous simulation of two output parameters, a multilayer ANN model was proposed as the best configuration. Therefore, an architecture of four neurons in the first hidden layer and four neurons in the second hidden layer (08:04:04:02) was sufficient to reproduce the output parameters, achieving a value of R2 of 0.9265, 0.9573 and with a mean absolute percentage error of 2.41, 1.14% for COP and COPCarnot, respectively. In the three configurations, the use of hyperbolic tangent sigmoid activation function (TANSIG) in the hidden layers and the adjustment of the coefficients with the Levenberg–Marquardt learning algorithm obtained the best results. The influence of each of the variables selected for the ANN model was analyzed through a correlation matrix and a sensitivity analysis. Other experimental variables were added in the training of the ANN model to consult the impact caused during the simultaneous prediction of the performance coefficients.
Similar content being viewed by others
References
International Energy Agency (2022) World Key energy statistics 2019. https://www.iea.org/world. Accessed 19 April 2022
Peixer GF, Dutra SL, Calomeno RS, SA NM, Lang GB, Lozano JA, Barbosa JJR (2022) Influence of heat exchanger design on the thermal performance of a domestic wine cooler driven by a magnetic refrigeration system. An Acad Bras Ciênc. https://doi.org/10.1590/0001-3765202220200563
Cudok F, Giannetti N, Ciganda JLC, Aoyama J, Babu P, Coronas A, Fujii T, Inoue N, Saito K, Yamaguchi S, Ziegler F (2021) Absorption heat transformer - state-of-the-art of industrial applications. Renew Sust Energ Rev 141:110757. https://doi.org/10.1016/j.rser.2021.110757
Alonso D, Cachot T, Hornut JM (2003) Experimental study of an innovative absorption heat transformer using partially miscible working mixtures. Int J Therm Sci 42:631–638. https://doi.org/10.1016/S1290-0729(03)00028-0
Rivera W, Huicochea A, Romero RJ, Lozano A (2018) Experimental assessment of double-absorption heat transformer operating with H2O/LiBr. Appl Therm Eng 132:432–440. https://doi.org/10.1016/j.applthermaleng.2017.12.117
Saito K, Inoue N, Nakagawa Y, Fukusumi Y, Yamada H, Irie T (2015) Experimental and numerical performance evaluation of double-lift absorption heat transformer. Sci Technol Built Environ 21:312–322. https://doi.org/10.1080/23744731.2014.998937
Cudok F, Ciganda JLC, Kononenko N, Drescher E (2017) Experimental results of an absorption heat transformer. In; 12th IEA Heat pump conference, Netherlands
Sharma VK, Kumar EA (2018) Thermodynamic simulation of hydrogen based solid sorption heat transformer. Int J Therm Sci 125:74–80. https://doi.org/10.1016/j.ijthermalsci.2017.11.020
Gómez-Arias E, Ibarra-Bahena J, Velazquez-Avelar L, Romero RJ, Rodríguez-Martínez A, Montiel-González M (2014) Temperature and concentration fields in a generator integrated to single stage heat transformer using Water/Carrol mixture. J Therm Sci 23:564–571. https://doi.org/10.1007/s11630-014-0742-2
Chaiyat N, Kiatsiriroat T (2014) Simulation and experimental study of solar-absorption heat transformer integrating with two-stage high temperature vapor compression heat pump. Case Stud Therm Eng 4:166–174. https://doi.org/10.1016/j.csite.2014.09.007
Yari M, Salehi S, Mahmoudi SMS (2017) Three-objective optimization of water desalination systems based on the double-stage absorption heat transformers. Desalination 405:10–28. https://doi.org/10.1016/j.desal.2016.12.001
Xu X, Liu J, Wang Y, Xu J, Bao J (2020) Performance evaluation of ground source heat pump using linear and nonlinear regressions and artificial neural networks. Appl Therm Eng 180:115914. https://doi.org/10.1016/j.applthermaleng.2020.115914
Zhang N, Zou H, Zhang L, Puppala AJ, Liu S, Cai G (2020) A unified soil thermal conductivity model based on artificial neural network. Int J Therm Sci 155:106414. https://doi.org/10.1016/j.ijthermalsci.2020.106414
Wang X, Yan X, Gao N, Chen G (2020) Prediction of thermal conductivity of various nanofluids with ethylene glycol using artificial neural network. J Therm Sci 29:1504–1512. https://doi.org/10.1007/s11630-019-1158-9
Moustafa EB, Hammad AH, Elsheikh AH (2022) A new optimized artificial neural network model to predict thermal efficiency and water yield of tubular solar still. Case Stud Therm Eng 30:101750. https://doi.org/10.1016/j.csite.2021.101750
Mitra A, Majumdar A, Majumdar PK, Bannerjee D (2013) Predicting thermal resistance of cotton fabrics by artificial neural network model. Exp Therm Fluid Sci 50:172–177. https://doi.org/10.1016/j.expthermflusci.2013.06.006
Corrales JLC, Cudok F (2020) Steady state measurements of an absorption heat transformer prototype. Repository. Technische Universität Berlin
Morales LI, Juárez-Romero D, Siqueiros J, Torres-Merino J (2013) Pre-startup proofs in an absorption heat transformer for water purification with compact components. In: ISES Solar World Congress, Energy Procedia
Cengel YA, Cimbala JM (2006) Fluid mechanics: fundamentals and applications, 1st edn. McGraw-Hill, New York
Abrahamsson K, Jernqvist Å (1993) Carnot comparison of multi-temperature level absorption heat cycles. Int J Refrig 16:240–246. https://doi.org/10.1016/0140-7007(93)90076-K
Horuz I, Kurt B (2010) Absorption heat transformers and an industrial application. Renew Energy 35:2175–2181. https://doi.org/10.1016/j.renene.2010.02.025
Mohanraj M, Jayaraj S, Muraleedharan C (2015) Applications of artificial neural networks for thermal analysis of heat exchangers: a review. Int J Therm Sci 90:150–172. https://doi.org/10.1016/j.ijthermalsci.2014.11.030
Beale MH, Hagan MT, Demuth HB (1992) Neural network toolbox user’s guide, vol 103. The MathWorks Inc
Feng YQ, Liu YZ, Wang X, He ZX, Hung TC, Wang Q, Xi H (2020) Performance prediction and optimization of an organic Rankine cycle (ORC) for waste heat recovery using back propagation neural network. Energy Convers Manag 226:113552. https://doi.org/10.1016/j.enconman.2020.113552
Sheskin DJ (2000) Handbook of parametric and nonparametric statistical procedures, 2nd edn. CHAPMAN & HALL/CRC, Boca Raton
Lind DA, Mason RD, Marchal WG (2001) Estadística para administración y economía, 3a edn. Irwin McGraw Hill
Martinez-Martinez E, Escobedo-Trujillo AB, Colorado D, Morales LI, Huicochea A, Hernández JA, Siqueiros J (2017) Criteria for improving the traditional artificial neural network methodology applied to predict COP for a heat transformer. Desalin Water Treat 73:90–100. https://doi.org/10.5004/dwt.2017.20357
Fan YV, Klemeš JJ, Lee CT, Tan RR (2021) Demographic and socio-economic factors including sustainability related indexes in waste generation and recovery. Energy Sourc Part A: Recover Utilization Environ Eff. https://doi.org/10.1080/15567036.2021.1974610
Garson GD (1991) Interpreting neural-network connection weights. AI Expert 6:47–51
Esfe MH, Afrand M, Wongwises S, Naderi A, Asadi A, Rostami S, Akbari M (2015) Applications of feedforward multilayer perceptron artificial neural networks and empirical correlation for prediction of thermal conductivity of Mg(OH)2–EG using experimental data. Int Commun Heat Mass Transf 67:46–50. https://doi.org/10.1016/j.icheatmasstransfer.2015.06.015
Dhande DY, Choudhari CS, Gaikwad DP, Dahe KB (2022) Development of artificial neural network to predict the performance of spark ignition engine fuelled with waste pomegranate ethanol blends. Inf Process Agric. https://doi.org/10.1016/j.inpa.2022.05.001
Yaïci W, Entchev E (2014) Performance prediction of a solar thermal energy system using artificial neural networks. Appl Therm Eng 73:1348–1359. https://doi.org/10.1016/j.applthermaleng.2014.07.040
Song J, Chen Y, Wu X, Ruan S, Zhang Z (2021) A novel approach for energy efficiency prediction of various natural draft wet cooling towers using ANN. J Therm Sci 30:859–868. https://doi.org/10.1007/s11630-020-1296-0
Muralidharan K, Vasudevan D (2015) Applications of artificial neural networks in prediction of performance, emission and combustion characteristics of variable compression ratio engine fuelled with waste cooking oil biodiesel. J Braz Soc Mech Sci Eng 37:915–928. https://doi.org/10.1007/s40430-014-0213-4
Acknowledgements
The corresponding authors thank CONACYT-SNI for the support provided. All the members involved in this work thank the authors mentioned in [7] for sharing the experimental data, collected from the repository of the Technische Universität Berlin.
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Technical Editor: Guilherme Ribeiro.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix 1
Weighting and bias coefficients produced from the ANN model for the prediction of the COP.
Number of neurons (s) | \({\text{Wi}}\) | \({\text{Wo}}\) | Bias | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Variable | Output layer (l = 1) | ||||||||||
\(T_{1}\) (k = 1) | \(T_{2}\) (k = 2) | \(T_{3}\) (k = 3) | \(T_{4}\) (k = 4) | \(T_{5}\) (k = 5) | \(T_{6}\) (k = 6) | \(T_{7}\) (k = 7) | \(T_{8}\) (k = 8) | \(Wo_{1,s}\) | \(b1_{\left( s \right)}\) | \(b2_{\left( l \right)}\) | |
1 | − 2.1331 | − 0.4083 | − 0.7737 | 0.7099 | 2.3708 | 2.1206 | 0.7128 | 1.8999 | 0.0181 | 0.6232 | 0.5970 |
2 | 0.2441 | − 1.3668 | − 2.0000 | 2.0153 | − 0.9345 | − 1.8978 | − 0.4147 | 2.3316 | 0.0196 | − 0.6881 | |
3 | − 0.9422 | 0.0293 | 0.3891 | 0.4074 | − 2.9774 | 1.0543 | 2.4282 | 1.5277 | − 0.0635 | − 0.9256 | |
4 | − 1.7725 | − 2.3908 | 0.7713 | 1.4164 | 2.0856 | − 1.3312 | 1.3290 | 0.1880 | − 0.2312 | − 0.3950 | |
5 | 1.1707 | 0.5308 | 2.0106 | 1.5114 | 1.0232 | 0.6669 | − 0.6753 | − 3.2822 | 0.0064 | − 1.3759 | |
6 | 0.1900 | − 2.7516 | 1.1800 | 0.9036 | − 1.7018 | 0.8779 | 0.9966 | − 1.9703 | 0.0056 | 2.8516 | |
7 | − 1.6872 | 1.7982 | 1.1574 | 1.9072 | − 0.2321 | − 0.4701 | 0.9764 | 2.1916 | 0.0681 | − 5.6599 |
Weighting and bias coefficients produced from the ANN model for the prediction of the COPCarnot.
\({\text{Wi}}\) | \({\text{Wo}}\) | Bias | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Number of neurons (s) | Variable | Output layer (l = 1) | |||||||||
\(T_{1}\) (k = 1) | \(T_{2}\) (k = 2) | \(T_{3}\) (k = 3) | \(T_{4}\) (k = 4) | \(T_{5}\) (k = 5) | \(T_{6}\) (k = 6) | \(T_{7}\) (k = 7) | \(T_{8}\) (k = 8) | \(Wo_{1,s}\) | \(b1_{\left( s \right)}\) | \(b2_{\left( l \right)}\) | |
1 | 1.2221 | 0.9570 | − 2.1169 | − 1.1600 | 1.2064 | 2.6072 | − 0.8197 | − 0.7379 | − 0.1288 | − 2.2468 | − 0.1578 |
2 | − 1.0986 | 2.3541 | 1.9413 | − 1.0455 | 1.7354 | − 0.9605 | − 0.5717 | 2.1373 | − 0.2415 | − 0.6347 | |
3 | − 2.2385 | − 1.0154 | 0.2286 | − 2.0534 | 0.9983 | 2.0025 | − 1.0321 | − 1.7013 | 0.0497 | 3.4225 | |
4 | − 0.3122 | 0.0743 | 2.4414 | 1.1720 | − 1.8615 | − 0.6122 | − 0.8637 | 2.5789 | 0.0107 | − 1.2252 | |
5 | − 2.3927 | 1.7122 | 1.3911 | 1.7826 | − 0.5359 | − 1.7792 | − 1.7396 | 0.7528 | − 0.6157 | − 1.3462 | |
6 | − 1.7680 | − 1.6591 | 1.5389 | − 0.5904 | 1.2875 | − 2.0185 | 0.1877 | 2.2252 | − 0.3378 | − 0.0802 | |
7 | − 0.1275 | 1.5380 | − 2.9774 | − 2.2634 | 0.2984 | 1.4155 | 1.9892 | 0.2671 | 0.1017 | − 2.2307 |
Appendix 2
Weighting and bias coefficients produced from the ANN model for the simultaneous prediction of COP and COPCarnot.
\({\text{Wi}}\) | Wo | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Number of neurons (s) | Variable | Output layer, (\(l = 1\)) | Output layer, (\(l = 2\)) | ||||||||||
\(T_{1}\) (k = 1) | \(T_{2}\) (k = 2) | \(T_{3}\) (k = 3) | \(T_{4}\) (k = 4) | \(T_{5}\) (k = 5) | \(T_{6}\) (k = 6) | \(T_{7}\) (k = 7) | \(T_{8}\) (k = 8) | \(b_{1\left( s \right)}\) | \({\text{Wo}}_{1,s}\) | \(b_{{2\left( {1,{\text{s}}} \right)}}\) | \({\text{Wo}}_{2,s}\) | \(b_{{2\left( {2,s} \right)}}\) | |
1 | 1.6747 | − 0.5274 | − 1.2592 | 2.0081 | 1.3907 | − 2.5409 | − 1.7215 | − 2.2429 | − 0.7503 | 0.8681 | − 0.1336 | 0.2726 | − 0.3537 |
2 | − 1.9968 | 1.3322 | 0.3864 | 0.9460 | 0.6943 | − 2.8574 | − 0.0128 | 2.9998 | 1.6963 | 0.5568 | 0.3913 | ||
3 | 1.0649 | 1.8916 | 0.5598 | 1.1567 | − 2.6749 | 0.7416 | − 0.9553 | − 3.1411 | − 0.8227 | − 0.0196 | 0.2331 | ||
4 | − 2.9780 | 0.0266 | − 1.4290 | − 0.3328 | − 1.3324 | − 1.6011 | 1.3466 | − 2.1245 | 5.9750 | 0.0486 | − 0.0470 | ||
5 | 3.1788 | − 0.7218 | − 1.8047 | 0.4607 | 0.2179 | − 2.1842 | − 1.2756 | − 2.3456 | 1.5401 | − 0.1880 | 0.2781 | ||
6 | 1.7075 | 0.4295 | 2.1427 | − 2.0244 | 0.5295 | 1.1132 | 2.6168 | 2.4397 | − 4.9063 | − 0.0085 | 0.1543 | ||
7 | − 1.2740 | − 2.8032 | 0.3272 | 1.2777 | − 1.1816 | − 1.7473 | − 2.1240 | 2.4022 | 2.4572 | − 0.0075 | − 0.0177 | ||
8 | 2.5366 | − 1.4103 | 1.1171 | 2.7357 | − 1.5617 | − 1.1458 | − 0.7763 | − 1.2546 | − 0.8566 | − 0.1683 | − 0.0054 | ||
9 | 2.3108 | 0.3890 | 2.5240 | − 1.3345 | 0.0620 | − 2.5060 | 1.8813 | 0.4995 | − 1.3727 | − 0.3583 | − 0.1582 | ||
10 | − 1.2346 | 1.5734 | 2.6753 | − 2.4446 | 0.2297 | 1.2647 | 1.8880 | − 0.8900 | − 2.6726 | 0.3199 | 0.0300 | ||
11 | 1.7013 | − 2.2208 | − 0.4385 | − 2.1839 | 0.2297 | − 1.4996 | − 2.2048 | 0.3056 | 3.1038 | 1.1650 | 0.1081 | ||
12 | − 1.0138 | 2.5953 | − 2.7922 | 1.6915 | − 0.9334 | − 1.9297 | − 1.3175 | − 1.0216 | 0.6049 | − 0.0304 | − 0.4375 | ||
13 | − 1.5880 | − 0.6965 | − 0.7669 | 3.1960 | − 2.2016 | 1.6127 | − 1.7808 | 0.5850 | − 0.7899 | 0.4847 | − 0.1436 | ||
14 | 0.4511 | 2.7091 | − 0.2609 | − 0.3156 | 2.2022 | − 2.4428 | − 1.6971 | 1.1723 | 0.0231 | 0.6328 | 0.1326 | ||
15 | − 1.1389 | 2.1478 | 2.4430 | 0.6585 | − 2.5051 | 0.9171 | − 0.9272 | 2.4435 | − 3.0446 | 1.0608 | 0.0156 | ||
16 | 1.8798 | − 2.5119 | 2.3131 | − 2.5288 | 1.2460 | − 0.1742 | 0.3855 | 1.1100 | 1.4270 | 0.8586 | 0.6079 |
Appendix 3
Weighting and bias coefficients produced from the multilayer ANN model for the simultaneous prediction of COP and COPCarnot.
(s1) | \({\text{Wi}}\) | \({\text{Wo}}\) | LW | ||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
First hidden layer | Bias | (s2) | Second hidden layer | Bias | Output layers | Bias | |||||||||||||
\(T_{1}\) (k = 1) | \(T_{2}\)(k = 2) | \(T_{3}\) (k = 3) | \(T_{4}\) (k = 4) | \(T_{5}\) (k = 5) | \(T_{6}\) (k = 6) | \(T_{7}\) (k = 7) | \(T_{8}\) (k = 8) | \(b_{{1\left( {\text{s}} \right)}}\) | \(s_{1} = 1\) | \(s_{1} = 2\) | \(s_{1} = 3\) | \(s_{1} = 4\) | \(b_{{2\left( {{\text{s}}2} \right)}}\) | \({\text{LW}}_{s3,1}\) | \({\text{LW}}_{s3,2}\) | \(b_{{3\left( {{\text{s}}3,1} \right)}}\) | \(b_{{3\left( {{\text{s}}3,2} \right)}}\) | ||
1 | 1.9887 | 0.7834 | 1.0414 | − 1.7030 | 1.0367 | 1.2334 | − 1.5372 | 1.9950 | − 3.5116 | 1 | 1.5136 | 1.2618 | 0.4001 | − 0.1371 | − 2.2623 | − 0.0821 | − 0.0838 | − 0.7702 | − 0.1314 |
2 | − 1.0213 | 0.2249 | 0.4104 | 0.7932 | 1.3645 | 1.9303 | 2.4361 | 1.6075 | − 4.5641 | 2 | − 1.0002 | 1.0356 | 0.6990 | − 1.0029 | 0.8045 | − 0.0059 | − 0.1786 | ||
3 | 0.0483 | − 2.0885 | − 1.3767 | 0.5722 | − 2.0218 | − 1.8054 | − 1.5816 | 0.7019 | 4.2296 | 3 | 0.4918 | − 0.2132 | 0.6851 | − 1.1641 | 1.2014 | − 0.0076 | 0.1381 | ||
4 | 2.0201 | − 0.0463 | − 2.3453 | 0.6314 | 0.2979 | 1.6527 | − 2.6486 | − 0.5006 | 0.3462 | 4 | 0.6300 | 1.0686 | 0.9701 | 1.3983 | 2.2163 | 1.1926 | 0.7475 |
Appendix 4
Weighting and bias coefficients produced from the ANN model contemplating twelve input variables for the simultaneous prediction of COP and COPCarnot.
Number of neurons (s) | \({\text{Wi}}\) | Wo | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Variable | Output layer, (\(l = 1\)) | Output layer, (\(l = 2\)) | |||||||||||||||
\(T_{1}\) (k = 1) | \(T_{2}\) (k = 2) | \(T_{3}\) (k = 3) | \(T_{4}\) (k = 4) | \(T_{5}\) (k = 5) | \(T_{6}\) (k = 6) | \(T_{7}\) (k = 7) | \(T_{8}\) (k = 8) | \(P_{1}\) (k = 9) | \(P_{2}\) (k = 10) | \(X_{1}\) (k = 11) | \(X_{2}\) (k = 12) | \(b_{1\left( s \right)}\) | \({\text{Wo}}_{1,s}\) | \(b_{{2\left( {1,{\text{s}}} \right)}}\) | \({\text{Wo}}_{2,s}\) | \(b_{{2\left( {2,s} \right)}}\) | |
1 | − 1.6138 | − 2.1650 | 0.7990 | − 1.4483 | 0.9342 | − 1.6123 | − 0.5266 | − 0.4265 | − 1.6099 | − 0.8184 | 1.9543 | 0.4207 | 4.6934 | 0.1277 | − 0.6248 | − 0.0595 | 0.0530 |
2 | 1.3654 | − 1.2615 | − 1.9379 | − 1.3412 | 1.8591 | − 0.6898 | 0.8984 | − 2.1796 | − 0.1650 | 0.3952 | − 0.3812 | 1.3437 | 0.8860 | 0.1213 | 0.0473 | ||
3 | 0.2644 | 1.2766 | 1.5811 | 1.1820 | 0.3581 | 0.0344 | − 0.5857 | − 1.8320 | − 0.4457 | 2.6837 | 1.0152 | 0.9165 | − 2.8019 | 0.2323 | − 0.0504 | ||
4 | − 0.3920 | − 1.7987 | 0.8456 | − 0.7068 | 0.7527 | − 0.5965 | − 1.3426 | 0.2267 | 2.4616 | 0.3208 | 1.1875 | 0.5110 | 0.6268 | 0.1933 | − 0.3198 | ||
5 | − 1.5465 | − 1.8394 | 0.5179 | 1.6199 | 0.9502 | − 0.3752 | 1.6878 | − 0.0382 | − 1.0893 | − 0.9167 | 1.6557 | − 1.0934 | − 1.7476 | − 0.9754 | − 0.8633 |
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Conde-Gutiérrez, R.A., Colorado, D., Gonzalez-Flores, P.B. et al. Simultaneous prediction of the performance coefficients in a compact absorption heat transformer using new neural network configurations. J Braz. Soc. Mech. Sci. Eng. 45, 426 (2023). https://doi.org/10.1007/s40430-023-04329-0
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s40430-023-04329-0