{"title":"Unified physics-informed ResNet model for comprehensive performance prediction of finned-tube evaporators","authors":"Bo Zhang, Xing-Yu Liang, Chun-Lu Zhang","doi":"10.1016/j.ijrefrig.2025.03.042","DOIUrl":null,"url":null,"abstract":"<div><div>Finned-tube heat exchangers are widely used in refrigeration and thermal engineering. With the growing demand for digital twins in complex systems, the need for fast and accurate performance prediction has become more urgent. Existing neural network models often suffer from fragmented structures, incomplete parameter considerations, and an inability to capture essential physical details. This study presents a unified physics-informed residual network model that overcomes these limitations by ensuring parameter completeness and integrating physical principles into the network architecture. The proposed method avoids redundant input parameters and conforms to physical laws. Moreover, it ensures that the selected output parameters meet the requirements of actual performance prediction and are conducive to the training of neural networks. By combining residual blocks with independent layers, the model achieves joint prediction of multiple performance metrics, improving both efficiency and accuracy. The proposed model significantly enhances predictive performance. All the determination coefficients reach 0.999, while both the mean absolute error and root mean square error values remain remarkably low. Specifically, it achieves a mean absolute percentage error of 0.41% for total heat transfer rate, 0.65% for sensible heat transfer rate, and below 0.3% for both refrigerant and air pressure drop. Furthermore, it effectively captures critical physical details, such as transitions between dry and wet operating conditions. Compared to previous models, this approach provides a more comprehensive, physically consistent, and computationally efficient framework for finned-tube evaporator performance prediction.</div></div>","PeriodicalId":14274,"journal":{"name":"International Journal of Refrigeration-revue Internationale Du Froid","volume":"175 ","pages":""},"PeriodicalIF":3.5000,"publicationDate":"2025-03-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Journal of Refrigeration-revue Internationale Du Froid","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0140700725001355","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, MECHANICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Finned-tube heat exchangers are widely used in refrigeration and thermal engineering. With the growing demand for digital twins in complex systems, the need for fast and accurate performance prediction has become more urgent. Existing neural network models often suffer from fragmented structures, incomplete parameter considerations, and an inability to capture essential physical details. This study presents a unified physics-informed residual network model that overcomes these limitations by ensuring parameter completeness and integrating physical principles into the network architecture. The proposed method avoids redundant input parameters and conforms to physical laws. Moreover, it ensures that the selected output parameters meet the requirements of actual performance prediction and are conducive to the training of neural networks. By combining residual blocks with independent layers, the model achieves joint prediction of multiple performance metrics, improving both efficiency and accuracy. The proposed model significantly enhances predictive performance. All the determination coefficients reach 0.999, while both the mean absolute error and root mean square error values remain remarkably low. Specifically, it achieves a mean absolute percentage error of 0.41% for total heat transfer rate, 0.65% for sensible heat transfer rate, and below 0.3% for both refrigerant and air pressure drop. Furthermore, it effectively captures critical physical details, such as transitions between dry and wet operating conditions. Compared to previous models, this approach provides a more comprehensive, physically consistent, and computationally efficient framework for finned-tube evaporator performance prediction.
期刊介绍:
The International Journal of Refrigeration is published for the International Institute of Refrigeration (IIR) by Elsevier. It is essential reading for all those wishing to keep abreast of research and industrial news in refrigeration, air conditioning and associated fields. This is particularly important in these times of rapid introduction of alternative refrigerants and the emergence of new technology. The journal has published special issues on alternative refrigerants and novel topics in the field of boiling, condensation, heat pumps, food refrigeration, carbon dioxide, ammonia, hydrocarbons, magnetic refrigeration at room temperature, sorptive cooling, phase change materials and slurries, ejector technology, compressors, and solar cooling.
As well as original research papers the International Journal of Refrigeration also includes review articles, papers presented at IIR conferences, short reports and letters describing preliminary results and experimental details, and letters to the Editor on recent areas of discussion and controversy. Other features include forthcoming events, conference reports and book reviews.
Papers are published in either English or French with the IIR news section in both languages.