Author(s)

Chao Si^*, Yuan Zhang, YanKun Li

Affiliation(s)

Electrical Department, China Institute of Atomic Energy, Beijing 102400, China.

Corresponding Author

Chao Si

ABSTRACT

As a typical ultra-large-scale, strongly coupled, and high-safety-level complex system engineering, nuclear power engineering involves multi-stage, multi-professional, and multi-stakeholder collaborative links throughout its full life cycle, including design, procurement, construction, commissioning, and operation and maintenance. Traditional management models generally face problems such as prominent information silos, inconsistent professional interfaces, high supply chain risks, lagging construction monitoring, and heavy pressure on quality and safety management, which are difficult to meet the requirements of current large-scale nuclear power construction and digital transformation. With the rapid development of artificial intelligence, big data, digital twin, BIM, and Internet of Things technologies, EPC full-process collaborative management driven by intelligent algorithms has gradually become an important development direction of nuclear power engineering management technology. This paper systematically analyzes the complexity characteristics of nuclear power engineering EPC, and constructs a full-process collaborative theoretical system of "data connection - scenario modeling - algorithm driving - platform support" from the dual perspectives of design institutes and owners. Focusing on intelligent design optimization, multi-professional collaboration, intelligent procurement management, construction progress identification and prediction, intelligent quality and safety monitoring, and operation and maintenance based on digital twins, a set of engineering implementable intelligent management schemes is proposed. Meanwhile, by constructing a cloud-edge-end integrated collaborative platform and MLOps model system, a technical architecture supporting full-life-cycle intelligent decision-making is formed. The verification results in typical nuclear power engineering projects show that artificial intelligence technology can increase design efficiency by more than 80%, reduce professional conflicts by more than 70%, lower procurement risks by 30%, decrease construction rework rate by more than 60%, and improve equipment availability to more than 98%, effectively supporting the quality improvement, cost reduction, and efficiency enhancement of nuclear power projects. The research results provide a systematic method and practical sample for the intelligent upgrading of nuclear power EPC, and have important theoretical significance and engineering value for promoting the digitalization and intelligent development of China's nuclear power industry.

KEYWORDS

Artificial intelligence; Nuclear power engineering; EPC management; Digital twin; Intelligent design; Intelligent operation and maintenance

CITE THIS PAPER

Chao Si, Yuan Zhang, YanKun Li. Artificial intelligence empowering nuclear power EPC full-process collaborative management. World Journal of Management Science. 2025, 3(5): 1-17. DOI: https://doi.org/10.61784/wms3092.

REFERENCES

[1] Farber J, Kobayashi K, Alajo A, et al. Integration of Control Methods and Digital Twins for Advanced Nuclear Reactors. Idaho National Laboratory (INL) / U.S. DOE, 2025.

[2] Kropaczek D J, Badalassi V, Jain P K, et al. Digital Twins for Nuclear Power Plants and Facilities// in Nuclear Engineering Handbook. Springer, 2023.

[3] U.S. Nuclear Regulatory Commission. NRC Research Activities for Nuclear Digital Twins. Washington, D.C.: NRC, 2024.

[4] Zhang X, Chen Y, Li Q, et al. Intelligent Design and Construction Coordination of Nuclear Power Based on Digital Twin// Proceedings of GEESD 2022. IOS Press, 2022: 276.

[5] Xing Y J, Li W Z, Zhang C D. Research on the Application of 3D Digital Twin Technology in Nuclear Power Engineering Construction. China Science and Technology Information, 2022(19): 45-50.

[6] Wu L, Liu Y S, Xue X, et al. Application of Grid Safety Management Model in Smart Construction Site of Nuclear Power Engineering. China Nuclear Power, 2024(6): 764–768.

[7] Do Y, Zebrowitz M, Stahl J, et al. iFANnpp: Nuclear power plant digital twin for robots and autonomous intelligence. Annals of Nuclear Energy, 2025: 111993.

[8] Daniell J, Kobayashi K, Alajo A, et al. Digital twin-centered hybrid data-driven multi-stage deep learning framework for enhanced nuclear reactor power prediction. arXiv:2211.13157, 2022.

[9] Hu M, Zhang X, Peng C, et al. Current status of digital twin architecture and application in nuclear energy field. Annals of Nuclear Energy, 2024: 110491.

[10] Hossain R B, Ahmed F, Kobayashi K, et al. Virtual sensing-enabled digital twin framework for real-time monitoring of nuclear systems leveraging deep neural operators. arXiv:2410.13762, 2024.