AI for Science in Next-Gen Power Systems: A Perspective

Wenxuan Liu; Junhua Zhao; Zhengping Lin; Zhao Yang Dong; Pierre Pinson

doi:10.1016/j.eng.2026.06.007

Engineering ›› :202606007 DOI: 10.1016/j.eng.2026.06.007

research-article

AI for Science in Next-Gen Power Systems: A Perspective

Author information +

History +

PDF (1198KB)

Abstract

In this perspective, we present a concise, visionary abstract that highlights the challenge of ever increasing complexity in future power grids and our proposed artificial intelligence for science (AI4S) framework. Next generation grids, with high levels of renewables and distributed generation, create unprecedented uncertainty that strains traditional control and forecasting methods. Our AI4S framework integrates cutting edge artificial intelligence (AI), including large language models (LLMs), physics-informed neural networks (PINNs), deep reinforcement learning (DRL), and autonomous multi-agent systems, within a domain aware architecture that embeds grid physics, operational constraints, and human oversight into learning. By fusing data driven models with power system laws, AI4S yields trustworthy, explainable AI that enables fast, real time decision making and richer situational awareness via digital twins and advanced forecasting. We distinguish which AI4S elements are already demonstrated versus conceptual, and we envision benefits like accelerated simulation, improved resilience, and greater operator trust.

Keywords

Artificial intelligence for science / Next-generation power systems / Physics-informed neural networks / Large language models / Digital twins

Cite this article

Download citation ▾

Wenxuan Liu, Junhua Zhao, Zhengping Lin, Zhao Yang Dong, Pierre Pinson. AI for Science in Next-Gen Power Systems: A Perspective. Engineering 202606007 DOI:10.1016/j.eng.2026.06.007

登录浏览全文

4963

注册一个新账户忘记密码

References

Publishing order | Descend order by publishing year | Descend order by cited within

[1]	Hou Q, Du E, Zhang N, Kang C . Impact of high renewable penetration on the power system operation mode: a data—driven approach. IEEE Trans Power Syst 2019; 35(1):731—41.

[2]	Saxena V, Manna S, Rajput SK, Kumar P, Sharma B, Alsharif MH, et al. Navigating the complexities of distributed generation: integration, challenges, and solutions. Energy Rep 2024; 12:3302—22.

[3]	Aien M, Hajebrahimi A, Fotuhi—Firuzabad M . A comprehensive review on uncertainty modeling techniques in power system studies. Renew Sustain Energy Rev 2016; 57:1077-89.

[4]	Dobbe R, Sondermeijer O, Fridovich—Keil D, Arnold D, Callaway D, Tomlin C . Toward distributed energy services: decentralizing optimal power flow with machine learning. IEEE Trans Smart Grid 2019; 11(2):1296-306.

[5]	Ahmadi M, Aly H, Gu J . A comprehensive review of AI—driven approaches for smart grid stability and reliability. Renew Sustain Energy Rev 2026; 226:116424.

[6]	Karniadakis GE, Kevrekidis IG, Lu L, Perdikaris P, Wang S, Yang L . Physics—informed machine learning. Nat Rev Phys 2021; 3(6):422—40.

[7]	Dogoulis P, Tit K, Cordy M . KCLNet: physics—informed power flow prediction via constraints projections. In: Proceedings of Machine Learning and Knowledge Discovery in Databases; 2025 Sep 15—19; Porto, Portugal. Cham: Springer; 2025. p. 95-110.

[8]	Cao Y, Zhao H, Cheng Y, Shu T, Chen Y, Liu G, et al. Survey on large language model—enhanced reinforcement learning: concept, taxonomy, and methods. IEEE Trans Neural Netw Learn Syst 2025; 36(6):9737—57.

[9]	Li Y, Arce A, Lucci T, Rasmussen RA, Lucks JB, Pei J, et al. Dynamic RNA synthetic biology: new principles, practices and potential. RNA Biol 2023; 20(1):817-29.

[10]	van der Velden BHM . Explainable AI: current status and future potential. Eur Radiol 2024; 34(2):1187-9.

[11]	Awais M, Naseer M, Khan S, Anwer RM, Cholakkal H, Shah M, et al. Foundation models defining a new era in vision: a survey and outlook. IEEE Trans Pattern Anal Mach Intell 2025; 47(4):2245-64.

[12]	Roumeliotis KI, Tselikas ND . ChatGPT and Open—AI models: a preliminary review. Future Int 2023; 15(6):192.

[13]	Huynh N, Lin B . Large language models for code generation: a comprehensive survey of challenges, techniques, evaluation, and applications. 2025. arXiv:2503.01245.

[14]	Alfarano A, Charton F, Hayat A . Global Lyapunov functions: a long—standing open problem in mathematics, with symbolic transformers. Adv Neural Inf Process Syst 2024; 37:93643-70.

[15]	Huang C, Li S, Liu R, Wang H, Chen Y . Large foundation models for power systems. In: Proceedings of 2024 IEEE Power & Energy Society General Meeting (PESGM); 2024 Jul 21—25; Seattle, WA, USA. IEEE; 2024.

[16]	Krishnan NAI. Agents: evolution, architecture, and real—world applications. 2025. arXiv:2503.12687.

[17]	Silver D, Schrittwieser J, Simonyan K, Antonoglou I, Huang A, Guez A, et al. Mastering the game of go without human knowledge. Nature 2017; 550(7676):354-9.

[18]	Gao S, Fang A, Huang Y, Giunchiglia V, Noori A, Schwarz JR, et al. Empowering biomedical discovery with AI agents. Cell 2024; 187(22):6125—51.

[19]	Angelopoulos A, Cahoon JF, Alterovitz R . Transforming science labs into automated factories of discovery. Sci Robot 2024; 9(95):eadm6991.

[20]	Ha T, Lee D, Kwon Y, Park MS, Lee S, Jang J, et al. AI—driven robotic chemist for autonomous synthesis of organic molecules. Sci Adv 2023; 9(44):eadj0461.

[21]	Darvish K, Skreta M, Zhao Y, Yoshikawa N, Som S, Bogdanovic M, et al. ORGANA: a robotic assistant for automated chemistry experimentation and characterization. Matter 2025; 8(2):101897.

[22]	Lu C, Lu C, Lange RT, Foerster J, Clune J, Ha D . The AI scientist: towards fully automated open—ended scientific discovery. 2024. arXiv:2408.06292.

[23]	Raissi M, Perdikaris P, Karniadakis GE . Physics—informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J Comput Phys 2019; 378:686-707.

[24]	Misyris GS, Venzke A, Chatzivasileiadis S . Physics—informed neural networks for power systems. In: Proceedings of 2020 IEEE Power & Energy Society General Meeting (PESGM); 2020 Aug 2—6; Montreal, QC, Canada. IEEE; 2020.

[25]	Arulkumaran K, Deisenroth MP, Brundage M, Bharath AA . Deep reinforcement learning: a brief survey. IEEE Signal Process Mag 2017; 34(6):26-38.

[26]	Degrave J, Felici F, Buchli J, Neunert M, Tracey B, Carpanese F, et al. Magnetic control of tokamak plasmas through deep reinforcement learning. Nature 2022; 602(7897):414—9.

[27]	Treloar NJ, Braniff N, Ingalls B, Barnes CP . Deep reinforcement learning for optimal experimental design in biology. PLOS Comput Biol 2022; 18(11):e1010695.

[28]

Chauhan A, Baranwal M, Basumatary A . PowRL: a reinforcement learning framework for robust management of power networks. In: Proceedings of the 37th AAAI Conference on Artificial Intelligence; 2023 Feb 7—14; Washington DC, USA. Association for the Advancement of Artificial Intelligence;2023. p. 14757—64.

[29]	Marot A, Donnot B, Dulac—Arnold G, Kelly A, O’Sullivan A, Viebahn J, et al. Learning to run a power network challenge: a retrospective analysis. In: Proceedings of NeurIPS 2020 Competition and Demonstration Track; 2020 Dec 6—12; online. PMLR; 2021. p. 112-32.

[30]	Marchesini E, Donnot B, Crozier C, Dytham I, Merz C, Schewe L, et al. RL2Grid: benchmarking reinforcement learning in power grid operations. 2025. arXiv:2503.23101.

[31]	Deihim A, Apostolopoulou D, Alonso E . Initial estimate of AC optimal power flow with graph neural networks. Electr Power Syst Res 2024; 234:110782.

[32]	Jiang Z, Lv H, Li Y, Guo Y . A novel application architecture of digital twin in smart grid. J Ambient Intell Humaniz Comput 2022; 13(8):3819-35.

[33]	Su T, Wu T, Zhao J, Scaglione A, Xie L . A review of safe reinforcement learning methods for modern power systems. 2024. arXiv:2407.00304.

[34]

Natarajan S, Mathur S, Sidheekh S, Stammer W, Kersting K . Human—in—the—loop or AI—in—the—loop? Automate or collaborate? In: Proceedings of the AAAI Conference on Artificial Intelligence; 2025 Feb 25—Mar 4; Philadelphia, Pennsylvania. Association for the Advancement of Artificial Intelligence; 2025. p. 28594-600.

[35]	Renugadevi R, Shobana J, Arthi K, Kalpana AV, Satishkumar D, Sivaraja M . Real—time applications of artificial intelligence technology in daily operations. In: Using real—time data and AI for thrust manufacturing. New York City: IGI Global Scientific Publishing; 2024. p. 243—57.

[36]	Liu G, Bai Y, Zhao H, Wen K, Wang X, Gu J, et al. MVLFLM: a parameter—efficient large language model framework for cross—domain multi—voltage load forecasting in smart grids. Energy AI 2025; 22:100651.

[37]	Wu X, Li YL, Sun J, Lu C . Symbol—LLM: leverage language models for symbolic system in visual human activity reasoning. Adv Neural Inf Process Syst 2023; 36:29680-91.

[38]	Cheng Y, Zhao H, Zhou X, Zhao J, Cao Y, Yang C, et al. A large language model for advanced power dispatch. Sci Rep 2025; 15(1):8925.

[39]	Nellikkath R, Chatzivasileiadis S . Physics—informed neural networks for AC optimal power flow. Electr Power Syst Res 2022; 212:108412.

[40]	Portu JC, Delle Femine C, Muro KS, Quartulli M, Restelli M . Limitations of physics—informed neural networks: a study on smart grid surrogation. In: Proceedings of 2025 IEEE Kiel PowerTech; 2025 Jun 29—Jul 3; Kiel, Germany. IEEE; 2025.

[41]	Xie W, Farazi F, Atherton J, Bai J, Mosbach S, Akroyd J, et al. Dynamic knowledge graph approach for modelling the decarbonization of power systems. Energy AI 2024; 17:100359.

[42]	Park S, Moura S, Lee K . Integration of hardware and software for battery hardware—in—the—loop toward battery artificial intelligence. IEEE Trans Transp Electrif 2024; 10(1):888-900.

[43]	Dewitte S, Cornelis JP, Müller R, Munteanu A . Artificial intelligence revolutionizes weather forecast, climate monitoring and decadal prediction. Remote Sens 2021; 13(16):3209.

[44]	Liu W, Zhao J, Qiu J, Dong ZY. Interpretable hybrid experimental learning for trading behavior modeling in electricity market. IEEE Trans Power Syst 2022; 38(2):1022—32.

[45]	Liu W, He B, Xue Y, Huang J, Zhao J, Wen F . A comprehensive modeling framework for coupled electricity and carbon markets. Energy Convers Econ 2024; 5(1):1-14.

[46]	Liu W, Zhao J, Huang J, Wen F, Lyu Z, Xue Y . Simulation and modelling analysis on behavior of cyber physical social system in energy: a case in electricity spot market. Autom Electr Power Syst 2020; 44(4):8-15. Chinese.

[47]	Malashin I, Tynchenko V, Gantimurov A, Nelyub V, Borodulin A . Physics—informed neural networks in polymers: a review. Polymers 2025; 17(8):1108.

[48]	Arévalo P, Cano A, Benavides J, Jurado F . Feasibility study of a renewable system (PV/HKT/GB) for hybrid tramway based on fuel cell and super capacitor. IET Renew Power Gener 2021; 15(3):491-503.

[49]	Bi K, Xie L, Zhang H, Chen X, Gu X, Tian Q . Accurate medium—range global weather forecasting with 3D neural networks. Nature 2023; 619(7970):533—8.

[50]	Bonavita M . On some limitations of current machine learning weather prediction models. Geophys Res Lett 2024; 51(12):e2023GL107377.

[51]	Brenowitz ND, Ge T, Subramaniam A, Gupta A, Hall DM, Mardani M, et al. Climate in a bottle: towards a generative foundation model for the kilometer—scale global atmosphere. 2025. arXiv:2505.06474.

[52]	Ali Z, Saad S, Rasheed K, Ammad S . AI Future Perspectives and Trends in Construction. In: AI in Material Science. CRC Press; 2024. p. 239-61.

[53]	Schneider T, Leung LR, Wills RC . Opinion: optimizing climate models with process knowledge, resolution, and artificial intelligence. Atmos Chem Phys 2024; 24(12):7041-62.

[54]	Materia S, García LP, van Straaten COS, Mamalakis A, Cavicchia L, Coumou D, et al. Artificial intelligence for climate prediction of extremes: state of the art, challenges, and future perspectives. Wiley Interdiscip Rev Clim Change 2024; 15(6):e914.

[55]	Song H, Liu CC, Lawarrée J . Nash equilibrium bidding strategies in a bilateral electricity market. IEEE Trans Power Syst 2002; 17(1):73-9.

[56]	Zhao H, Zhao J, Qiu J, Liang G, Wen F, Xue Y, et al. Data—driven risk preference analysis in day—ahead electricity market. IEEE Trans Smart Grid 2020; 12(3):2508—17.

[57]	Liang G, Liu G, Zhao J, Liu Y, Gu J, Sun G, et al. Super resolution perception for improving data completeness in smart grid state estimation. Engineering 2020; 6(7):789-800.

[58]	Wang X, Feng M, Qiu J, Gu J, Zhao J . From news to forecast: integrating event analysis in LLM—based time series forecasting with reflection. Adv Neural Inf Process Syst 2024; 37:58118-53.

[59]	Yu L, Wang P, Chen Z, Li D, Li N, Cherkaoui R . Finding Nash equilibrium based on reinforcement learning for bidding strategy and distributed algorithm for ISO in imperfect electricity market. Appl Energy 2023; 350:121704.

[60]	Naseri N, Talari S, Ketter W, Collins J . Dynamic retail market tariff design for an electricity aggregator using reinforcement learning. Electr Power Syst Res 2022; 212:108560.

[61]	Monti A, Dolley B, Palensky P, Parisot A . Building a digital twin for European Energy Infrastructure: the role of open source. IEEE Electrif Mag 2024; 12(3):78-84.

[62]	Saxena D, Singh AKA . Self—healing and fault—tolerant cloud—based digital twin processing management model. IEEE Trans Industr Inform 2025; 21(5):4233-42.

[63]	Zhao H, Zhao J, Liu W, Yan Y, Huang J, Wen F . Metaverse—based decentralised autonomous organisation in energy systems. Energy Convers Econ 2023; 4(6):379-86.