Physics-informed neural network-based digital twins for thermal energy systems: A review of solvability and loss function design

Abstract

Thermal energy systems (TES) are an essential part of industries that have evolved over time through the engagement of managers and researchers. The development of digital twin (DT) technology has enabled accurate prediction of their performance. The inherent limitations of complex thermal systems, such as noisy input data and occasional lack of measurement data or boundary conditions, have recently created opportunities to apply physics-based problem-solving alongside DT technology. This paper aims to systematically review the novel physics-informed neural network-digital twin (PINN-DT) methodology as a potential solution to these challenges, and to present a taxonomy for problem-solving. The outcome of this study provides valuable guidance in selecting PINN-DT technology in thermal energy system (TES) modeling. A review of the proposed loss functions demonstrates that their design is critical for achieving precise outcomes in this technology, effectively serving as the foundational core of PINN-DT. As a result, it is recommended that the construction of the loss function be fundamentally guided by two principal considerations: forecasting accuracy and compliance with physical principles, which serve as foundational pillars in the surrogate model design framework. A significant gap exists in applying this technology to industries that use discrete sampling for quality control. Implementing the PINN-DT framework could address this issue by determining optimal sampling intervals, thereby offering vital decision-making support. Moreover, the absence of exergy analysis in formulating the physical loss component of the loss function represents a significant research gap. Future studies should therefore incorporate the exergy concept into the design of the loss function.

Graphical abstract

Keywords

Physics-informed neural network-digital twin (PINN-DT) / Loss functions / Thermal energy system / Exergy analysis / Physics-informed loss functions

References

[1] Zhu B Y , Ren S J , Weng Q H . et al. A physics-informed variational autoencoder for modeling power plant thermal systems. Energies, 2025, 18(17): 4742

[2] Chen C , Liu M , Li M J . et al. Digital twin modeling and operation optimization of the steam turbine system of thermal power plants. Energy, 2024, 290: 129969

[3] Osorio J D , De Florio M , Hovsapian R . et al. Physics-Informed machine learning for solar-thermal power systems. Energy Conversion and Management, 2025, 327: 119542

[4] Das O , Zafar M H , Sanfilippo F . et al. Advancements in digital twin technology and machine learning for energy systems: A comprehensive review of applications in smart grids, renewable energy, and electric vehicle optimisation. Energy Conversion and Management: X, 2024, 24: 100715

[5] Tao F , Zhang H , Liu A . et al. Digital twin in industry: State-of-the-art. IEEE Transactions on Industrial Informatics, 2019, 15(4): 2405–2415

[6] Glaessgen E , Stargel D . The digital twin paradigm for future NASA and U.S. Air Force vehicles. In: Proceedings of the 53rd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, Honolulu, USA, 2012, 1818:

[7] Cimino C , Negri E , Fumagalli L . Review of digital twin applications in manufacturing. Computers in Industry, 2019, 113: 103130

[8] Farhat H , Altarawneh A . Physics-informed machine learning for intelligent gas turbine digital twins: A review. Energies, 2025, 18(20): 5523

[9] Raissi M , Perdikaris P , Karniadakis G E . Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational Physics, 2019, 378: 686–707

[10] Bendaouia A , Wang F F , Ouifak H . et al. Digital twin for battery energy storage systems. Renewable and Sustainable Energy Reviews, 2026, 226: 116347

[11] Rasheed A , San O , Kvamsdal T . Digital twin: Values, challenges and enablers from a modeling perspective. IEEE Access, 2020, 8: 21980–22012

[12] Liu M N , Fang S L , Dong H Y . et al. Review of digital twin about concepts, technologies, and industrial applications. Journal of Manufacturing Systems, 2021, 58: 346–361

[13] Karniadakis G E , Kevrekidis I G , Lu L . et al. Physics-informed machine learning. Nature Reviews Physics, 2021, 3(6): 422–440

[14] Jiang D Y , Wang Z L , Yuan L Q . et al. Physics-informed neural network for rapid prediction of the temperature fields in fuel-heat pipe assemblies. Energy, 2025, 332: 137268

[15] Chen T S , Kang Y J , Yan P B . et al. Supercritical carbon dioxide critical flow model based on a physics-informed neural network. Energy, 2024, 313: 133863

[16] Xu Y W , Kohtz S , Boakye J . et al. Physics-informed machine learning for reliability and systems safety applications: State of the art and challenges. Reliability Engineering & System Safety, 2023, 230: 108900

[17] Sun L N , Gao H , Pan S W . et al. Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data. Computer Methods in Applied Mechanics and Engineering, 2020, 361: 112732

[18] Wang Q B , Zhang X H , Lu J W . et al. Digital twin-driven physically constrained generative adversarial network for industrial boiler fault diagnosis. IEEE Transactions on Instrumentation and Measurement, 2025, 74: 3530215

[19] Chen D , Chui C K , Lee P S . Adaptive physically consistent neural networks for data center thermal dynamics modeling. Applied Energy, 2025, 377: 124637

[20] Li A , Guan F J , Hu H F . et al. Physics-informed DEEPLSTM for digital twin modeling of steam turbine full-operating condition. Applied Thermal Engineering, 2025, 278: 127439

[21] Xu B , Wang J N , Wang X P . et al. A case study of digital-twin-modelling analysis on power-plant-performance optimizations. Clean Energy, 2019, 3(3): 227–234

[22] Deon B , Cotta K P , Silva R F V . et al. Digital twin and machine learning for decision support in thermal power plant with combustion engines. Knowledge-Based Systems, 2022, 253: 109578

[23] Huang Z F , Soh K Y , Islam M R . et al. Digital twin driven life-cycle operation optimization for combined cooling heating and power-cold energy recovery (CCHP-CER) system. Applied Energy, 2022, 324: 119774

[24] Zhao Z , Stuebner M , Lua J . et al. Full-field temperature recovery during water quenching processes via physics-informed machine learning. Journal of Materials Processing Technology, 2022, 303: 117534

[25] Kannapinn M , Pham M K , Schäfer M . Physics-based digital twins for autonomous thermal food processing: Efficient, non-intrusive reduced-order modeling. Innovative Food Science & Emerging Technologies, 2022, 81: 103143

[26] Machado D O , Chicaiza W D , Escaño J M . et al. Digital twin of a Fresnel solar collector for solar cooling. Applied Energy, 2023, 339: 120944

[27] Ishitsuka K , Lin W R . Physics-informed neural network for inverse modeling of natural-state geothermal systems. Applied Energy, 2023, 337: 120855

[28] Zhang J C , Zhao X W . Digital twin of wind farms via physics-informed deep learning. Energy Conversion and Management, 2023, 293: 117507

[29] Baniasadi E , Rezk A , Tola Y B . et al. Renewable-driven hybrid refrigeration system for enhancing food preservation: Digital twin design and performance assessment. Energy Conversion and Management, 2024, 322: 119165

[30] Liang X Y , Zhang B , Zhang C L . Physics-informed deep residual neural network for finned-tube evaporator performance prediction. Energy, 2024, 302: 131822

[31] Lu Y K , Li Y F , Fu G C . et al. The physical information LSTM surrogate model for establishing a digital twin model of reciprocating air compressors. Applied Soft Computing, 2024, 167: 112309

[32] Wang T Y , Zhong W Q , Chen X . et al. A digital twin of coal-fired boiler for physical fields prediction using flow-field-informed POD-XGBoost reduced-order modeling. Energy, 2025, 329: 136732

[33] Ma M R , Zhang Z F , Chen J . et al. Physics-informed machine learning-driven active utilization of ventilation and heat exchange in large underground tunnels. Renewable Energy, 2025, 250: 123283

[34] Zerrougui I , Li Z L , Hissel D . Physics-Informed Neural Network for modeling and predicting temperature fluctuations in proton exchange membrane electrolysis. Energy and AI, 2025, 20: 100474

[35] Ye J D , Huang J X , Du J H . et al. Multi-objective optimization of supercritical carbon dioxide Brayton cycles using Bayesian physics-informed neural networks: A comprehensive analysis of energy, exergy, economic, and environmental performance. Energy, 2025, 339: 139033

[36] Han G , Hong S K , Moon J H . Physics-informed neural network digital twin for real-time energy-state tracking in multivariate thermochemical heat storage. Journal of Energy Storage, 2025, 139: 118686

[37] Garcia F A , Devriendt H , Metin H . et al. Physics-informed digital twin design for supporting the selection of process settings in continuous manufacturing, with a focus in fiberboard production. Computers in Industry, 2025, 168: 104267

[38] Ismail F B , Al-Faiz H , Hasini H . et al. A comprehensive review of the dynamic applications of the digital twin technology across diverse energy sectors. Energy Strategy Reviews, 2024, 52: 101334

[39] Liu X , Jiang D , Tao B . et al. A systematic review of digital twin about physical entities, virtual models, twin data, and applications. Advanced Engineering Informatics, 2023, 55: 101876

[40] Yu W , Patros P , Young B . et al. Energy digital twin technology for industrial energy management: Classification, challenges and future. Renewable and Sustainable Energy Reviews, 2022, 161: 112407

[41] Ghenai C , Husein L A , Al Nahlawi M . et al. Recent trends of digital twin technologies in the energy sector: A comprehensive review. Sustainable Energy Technologies and Assessments, 2022, 54: 102837

[42] Ataee S , Ameri M , Askari I B . et al. Evaluation and intelligent forecasting of energy and exergy efficiencies of a nanofluid-based filled-type U-pipe solar ETC using three machine learning approaches. Energy, 2024, 298: 131355

[43] Ataee S , Ameri M . Research on supercritical CO₂-based filled type U-pipe solar evacuated tubular collector: A thermodynamically parametric optimization and implementation machine learning modeling. Applied Thermal Engineering, 2025, 261: 125135

[44] Sleiti A K , Kapat J S , Vesely L . Digital twin in energy industry: Proposed robust digital twin for power plant and other complex capital-intensive large engineering systems. Energy Reports, 2022, 8: 3704–3726

[45] Lim J T , Habibullah A , Ng E Y K . Towards a digital twin for gas turbines: Thermodynamic modeling, critical parameter estimation, and performance optimization using PINN and PSO. Energies, 2025, 18(14): 3721

[46] Chen C , Fu H B , Zheng Y . et al. The advance of digital twin for predictive maintenance: The role and function of machine learning. Journal of Manufacturing Systems, 2023, 71: 581–594

[47] Liao S H , Jeong J , Zha R J . et al. Simulation-guided feedforward-feedback control of melt pool temperature in directed energy deposition. CIRP Annals, 2023, 72(1): 157–160

[48] Gunasegaram D R , Barnard A S , Matthews M J . et al. Machine learning-assisted in-situ adaptive strategies for the control of defects and anomalies in metal additive manufacturing. Additive Manufacturing, 2024, 8: 104013

[49] Chen W C , Liu H W , Qi E S . Discrete event-driven model predictive control for real-time work-in-process optimization in serial production systems. Journal of Manufacturing Systems, 2020, 55: 132–142

[50] Li X Y , Li N , Kolmanovsky I . et al. Stochastic model predictive control for remanufacturing system management. Journal of Manufacturing Systems, 2021, 59: 355–366

[51] Chen Y P , Karkaria V , Tsai Y K . et al. Real-time decision-making for Digital Twin in additive manufacturing with Model Predictive Control using time-series deep neural networks. Journal of Manufacturing Systems, 2025, 80: 412–424

[52] Ortner P , del Re L . Predictive control of a diesel engine air path. IEEE Transactions on Control Systems Technology, 2007, 15(3): 449–456

[53] Veysi P , Adeli M , Naziri N P. . Introduction to PID controller and model predictive control in engineering systems.. System, 2024, 1: 2

[54] Zobeiry N , Humfeld K D . A physics-informed machine learning approach for solving heat transfer equation in advanced manufacturing and engineering applications. Engineering Applications of Artificial Intelligence, 2021, 101: 104232

[55] Cai S Z , Wang Z C , Wang S F . et al. Physics-informed neural networks for heat transfer problems. Journal of Heat Transfer, 2021, 143(6): 060801

[56] Yan B C , Gudala M , Hoteit H . et al. Physics-informed machine learning for noniterative optimization in geothermal energy recovery. Applied Energy, 2024, 365: 123179

[57] You M L , Wang Q , Sun H J . et al. Digital twins based day-ahead integrated energy system scheduling under load and renewable energy uncertainties. Applied Energy, 2022, 305: 117899

[58] Mohammadi A , Sheikholeslam F . Intelligent optimization: Literature review and state-of-the-art algorithms (1965–2022). Engineering Applications of Artificial Intelligence, 2023, 126: 106959

[59] Liu Q , Leng J W , Yan D X . et al. Digital twin-based designing of the configuration, motion, control, and optimization model of a flow-type smart manufacturing system. Journal of Manufacturing Systems, 2021, 58: 52–64

[60] Raissi M , Yazdani A , Karniadakis G E . Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations. Science, 2020, 367(6481): 1026–1030

[61] Yang L , Meng X H , Karniadakis G E . B-PINNs: Bayesian physics-informed neural networks for forward and inverse PDE problems with noisy data. Journal of Computational Physics, 2021, 425: 109913

[62] Zhang J K , Wang Z T , Li S Y . et al. A digital twin approach for gas turbine performance based on deep multi-model fusion. Applied Thermal Engineering, 2024, 246: 122954

[63] Liu L L , Zhang X Y , Wan X . et al. Digital twin-driven surface roughness prediction and process parameter adaptive optimization. Advanced Engineering Informatics, 2022, 51: 101470

[64] Agostinelli S , Cumo F , Guidi G . et al. Cyber-physical systems improving building energy management: Digital twin and artificial intelligence. Energies, 2021, 14(8): 2338

[65] O’Dwyer E , Pan I , Charlesworth R . et al. Integration of an energy management tool and digital twin for coordination and control of multi-vector smart energy systems. Sustainable Cities and Society, 2020, 62: 102412

[66] Zhang Y B , Wang Z X , Filonenko K . et al. Physics-guided deep reinforcement learning for optimized data center cooling and waste heat recovery utilizing aquifer thermal energy storage. Applied Energy, 2026, 402: 126984

[67] Shokouhi P , Kumar V , Prathipati S . et al. Physics-informed deep learning for prediction of CO₂ storage site response. Journal of Contaminant Hydrology, 2021, 241: 103835

[68] Barimah A K , Onu O P , Niculita O . et al. Scalable data transformation models for physics-informed neural networks (PINNs) in digital twin-enabled prognostics and health management (PHM) applications. Computers, 2025, 14(4): 121

[69] Zhang S Y , Dinavahi V , Liang T . Towards hydrogen-powered electric aircraft: Physics-informed machine learning based multi-domain modeling and real-time digital twin emulation on FPGA. Energy, 2025, 322: 135451

[70] Khan B , Ali S M , Ullah Z . Deep learning based digital twins augmented reality: Model predictive control for battery and storage optimization in renewable energy prosumers districts. Journal of Energy Storage, 2025, 131: 117565

[71] Kingma D P , Ba J . Adam: A method for stochastic optimization. arXiv preprint: arXiv: 1412.6980, 2014,

[72] Byrd R H , Lu P H , Nocedal J . et al. A limited memory algorithm for bound constrained optimization. SIAM Journal on Scientific Computing, 1995, 16(5): 1190–1208

[73] Zhu T , Zheng Q Y , Lu Y L . Physics-informed fully convolutional networks for forward prediction of temperature field and inverse estimation of thermal diffusivity. Journal of Computing and Information Science in Engineering, 2024, 24(11): 111004

[74] Annamalai K , Puri I K , Jog M A . Advanced Thermodynamics Engineering. 2nd ed. Boca Raton: CRC Press, 2011,

[75] Maghsoudizadeh M , Ameri M , Jahanshahi Javaran E . Performance enhancement of the flash furnaces in copper industries via 4E analysis using hydrogen–natural gas blended fuel and oxygen-enriched air. Journal of Cleaner Production, 2025, 525: 146556

[76] Kazemi Naeini H , Shomali R , Pishahang A . et al. PINN-DT: Optimizing energy consumption in smart building using hybrid physics-informed neural networks and digital twin framework with blockchain security. Sensors, 2025, 25(19): 6242

[77] Wang X Z , Dong B . Physics-informed hierarchical data-driven predictive control for building HVAC systems to achieve energy and health nexus. Energy and Buildings, 2023, 291: 113088

[78] LoI K , Song C Y , Chao H P . et al. High temporal-resolution HVAC control in GPU-centric data centers via reinforcement learning and a knowledge-embedded physics-informed neural network. Journal of Building Engineering, 2026, 123: 115499

[79] Nguyen N T , Nguyen B , Nghiem T X . Physics-informed data-driven modeling of HVAC systems: A systematic analysis. IEEE Access, 2026, 14: 6481–6500

[80] Abedini Najafabadi H , Ozalp N , Ophoff C . et al. An experimental study on temperature control of a solar receiver under transient solar load. Solar Energy, 2019, 186: 52–59

[81] Nuntaphan A , Chansena C , Kiatsiriroat T . Performance analysis of solar water heater combined with heat pump using refrigerant mixture. Applied Energy, 2009, 86(5): 748–756

RIGHTS & PERMISSIONS

Higher Education Press