Journal of Nuclear Engineering & Technology Review Article
Digital Twin and Machine Learning Framework for Predictive Maintenance of Nuclear Power Systems
Abstract
Nuclear power systems contain safety-critical components whose degradation can affect availability, reliability, and maintenance planning. Conventional maintenance strategies are largely based on scheduled inspection, threshold monitoring, and condition-based assessment, which may not fully exploit the large volume of continuously generated plant data. Digital Twin (DT) technology provides a mechanism for maintaining a synchronized virtual representation of a physical nuclear system, while machine learning (ML) can extract degradation patterns from multivariate sensor measurements. This paper proposes a Digital Twin and Machine Learning Framework for Predictive Maintenance of Nuclear Power Systems. The framework combines reactor sensor data, physics-informed virtual modelling, data preprocessing, feature extraction, ML-based anomaly detection, fault classification, health estimation, and maintenance decision support. A representative multivariate reactor dataset containing 1,200 simulated operating samples and eight sensor variables is considered for methodological evaluation. Random Forest (RF), Support Vector Machine (SVM), Artificial Neural Network (ANN), and XGBoost models are investigated. Illustrative results indicate classification accuracies of 93.6%, 91.4%, 95.1%, and 97.2%, respectively, with XGBoost providing the highest simulated classification performance. A digital-twin layer continuously compares measured and predicted operating states to generate a health indicator and maintenance priority. Feature analysis identifies coolant flow, primary pressure, neutron flux, and coolant temperature as the dominant variables in the illustrative dataset. The proposed framework is designed as an operator-support and maintenance-planning system rather than a replacement for certified reactor protection systems. The study demonstrates how synchronized digital models and ML can provide an integrated foundation for predictive maintenance, anomaly diagnosis, and condition monitoring of nuclear power systems.
Keywords
References (11)
- Godase V, Pawar P, Nagane S, Kumbhar S. Automatic railway horn system using node MCU. Journal of Control & Instrumentation. 2024 Jan;15(1).
- Godase V, Godase J. Diet prediction and feature importance of gut microbiome using machine learning. Evolution in Electrical and Electronic Engineering. 2024 Nov 6;5(2):214-9.
- Jamadade VK, Ghodke MG, Katakdhond SS, Godase V. A comprehensive review on scalable Arduino radar platform for real-time object detection and mapping. Journal of Microprocessor and Microcontroller Research. 2025 May;2(2):1-2.
- Godase V. A comprehensive study of revolutionizing EV charging with solar-powered wireless solutions. Advance Research in Power Electronics and Devices e-ISSN. 2025 Apr 18:3048-7145.
- Rivas A, Delipei GK, Hou J. A system predictive maintenance framework for advanced reactors using a data-driven digital twin. Nuclear Science and Engineering. 2025 Mar 4;199(3):358-87.
- Fu Q, Wan Y, Liu Z. Reactor System Health Management Based on Deep Digital Twins. Nuclear Technology. 2025 Nov 24:1-4.
- Ayo-Imoru RM, Ali AA, Bokoro PN. An enhanced fault diagnosis in nuclear power plants for a digital twin framework. In2021 International Conference on Electrical, Computer and Energy Technologies (ICECET) 2021 Dec 9 (pp. 1-6). IEEE.
- Gancheva V. Digital Twin Framework for Nuclear Reactor Monitoring and Anomaly Detection. Energies. 2026 Jul 12;19(14):3278.
- Cancemi SA, Lo Frano R. The application of machine learning for on-line monitoring Nuclear Power Plant performance. InThe 30th International Conference Nuclear Energy for New Europe (NENE2021) 2021 Sep (pp. 1-9).
- Daniell J, Kobayashi K, Alajo A, Alam SB. Digital twin-centered hybrid data-driven multi-stage deep learning framework for enhanced nuclear reactor power prediction. Energy and AI. 2025 Jan 1;19:100450.
- Xu Q, Zhou G, Zhang C, Chang F, Huang Q, Zhang M, Zhi Y. A digital twin framework for nuclear power equipment maintenance: design, prototyping, and preliminary validation. Digital Twin. 2025 Jan 2;2(1):2482342.