Machine Learning Models for Predicting Microgrid Faults in Critical Healthcare Infrastructure
DOI:
https://doi.org/10.5281/zenodo.22644280Keywords:
microgrid fault detection, machine learning fault classification, random forest and ensemble learning, hospital backup power reliability, critical healthcare infrastructure resilienceAbstract
Background: Hospital and healthcare-facility microgrids increasingly combine distributed generation, battery storage, and
multiple simultaneously active sources to sustain life-critical electrical loads during utility outages, but this added architectural
complexity makes fault behavior difficult to detect and clear using conventional overcurrent-based protection. Machine learning
(ML) has become the leading approach to microgrid fault detection, classification, and prediction, yet much of the literature and
prior reviews emphasize artificial neural network and deep learning methods, leaving the broader landscape of classical and
ensemble ML models comparatively underexamined for critical-healthcare relevance. Objective: This narrative review synthesizes
published literature on classical machine learning models-including support vector machines, decision trees, random forests,
k-nearest neighbours, naive Bayes classifiers, and gradient-boosted ensembles such as XGBoost and LightGBM-alongside neural
network and deep learning approaches for microgrid fault detection, classification, and prediction, with particular attention to the
reliability and interpretability requirements of hospital and critical-healthcare infrastructure. Methods: A narrative synthesis
approach was used, drawing on peer-reviewed engineering studies of machine learning-based microgrid and power-system fault
diagnosis and on the separate literature addressing hospital microgrid design, resilience quantification, and healthcare-facility
infrastructural resilience, without a systematic meta-analytic pooling of effect sizes. Findings: Across the studies reviewed, classical
and ensemble ML models frequently matched or exceeded neural network and deep learning baselines on the same or comparable
microgrid and power-system fault-diagnosis tasks, with optimized random forest and heterogeneous stacked ensembles reaching
test accuracies above 99% for fault detection and classification while degrading only marginally under injected measurement
noise, and doing so with lower computational and training-data requirements and more auditable decision logic than many deep
architectures. Separately, the hospital microgrid resilience literature has developed detailed optimization and
resilience-quantification frameworks for sizing storage and generation against outage risk, but these frameworks generally treat
fault detection and clearing as an idealized, instantaneous background process rather than incorporating the finite detection
latency and residual misclassification risk that empirical ML models, classical or deep, actually exhibit. Conclusions: Classical and
ensemble machine learning models represent an underexploited but well-evidenced foundation for hospital-relevant microgrid
fault prediction, and integrating their documented accuracy, robustness, and interpretability characteristics into
hospital-representative resilience models is a necessary next step for translating laboratory-validated fault diagnosis into
dependable protection for critical healthcare infrastructure.
References
Hervas-Zaragoza J, Colmenar-Santos A, Rosales-Asensio E, Colmenar-Fernandez L (2022) Microgrids as a mechanism for improving energy resilience during grid outages: a post COVID-19 case study for hospitals. Renewable Energy, 199:308-319, DOI:10.1016/j.renene.2022.08.132.
Zidane TEK, Muis ZA, Ho WS, Zahraoui Y, Aziz AS, Su CL, Mekhilef S, Elia Campana P (2025) Power systems and microgrids resilience enhancement strategies: a review. Renewable and Sustainable Energy Reviews, 207:114953, DOI:10.1016/j.rser.2024.114953.
Zaben MM, Worku MY, Hassan MA, Abido MA (2024) Machine learning methods for fault diagnosis in AC microgrids: a systematic review. IEEE Access, 12:20260-20298, DOI:10.1109/ACCESS.2024.3360330.
Fahim SR, Sarker SK, Muyeen SM, Sheikh MRI, Das SK (2020) Microgrid fault detection and classification: machine learning based approach, comparison, and reviews. Energies, 13(13):3460, DOI:10.3390/en13133460.
Uzair M, Eskandari M, Li L, Zhu J (2022) Machine learning based protection scheme for low voltage AC microgrids. Energies, 15(24):9397, DOI:10.3390/en15249397.
Taheri A, Eskandari A, Heydari H (2026) Accurate diagnosis of line-to-ground and line-to-line-to-ground faults in AC microgrid with inverter-based distributed generators using random forest classifier. Computers and Electrical Engineering, 138:111317, DOI:10.1016/j.compeleceng.2026.111317.
Patnaik B, Mishra M, Bansal RC, Jena RK (2021) MODWT-XGBoost based smart energy solution for fault detection and classification in a smart microgrid. Applied Energy, 285:116457, DOI:10.1016/j.apenergy.2021.116457.
Vaish R, Dwivedi UD, Tewari S, Tripathi SM (2021) Machine learning applications in power system fault diagnosis: research advancements and perspectives. Engineering Applications of Artificial Intelligence, 106:104504, DOI:10.1016/j.engappai.2021.104504.
Saleh K, Ayad A (2021) Fault zone identification and phase selection for microgrids using decision trees ensemble. International Journal of Electrical Power & Energy Systems, 132:107178, DOI:10.1016/j.ijepes.2021.107178.
Lu Z, Wang L, Wang P (2024) Microgrid fault detection method based on lightweight gradient boosting machine-neural network combined modeling. Energies, 17(11):2699, DOI:10.3390/en17112699.
Anwar T, Mu C, Yousaf MZ, Khan W, Khalid S, Hourani AO, Zaitsev I (2025) Robust fault detection and classification in power transmission lines via ensemble machine learning models. Scientific Reports, 15:2549, DOI:10.1038/s41598-025-86554-2.
Najafzadeh M, Pouladi J, Daghigh A, Beiza J, Abedinzade T (2024) Fault detection, classification and localization along the power grid line using optimized machine learning algorithms. International Journal of Computational Intelligence Systems, 17:49, DOI:10.1007/s44196-024-00434-7.
Jayasinghe JARR, Malindi JHE, Rajapaksha RMAM, Logeeshan V, Wanigasekara C (2024) Classification and localization of faults in AC microgrids through discrete wavelet transform and artificial neural networks. IEEE Open Access Journal of Power and Energy, 11:303-313, DOI:10.1109/OAJPE.2024.3422387.
Esmaeilbeigi S, Kazemi Karegar H, Najar A (2025) A fault classification scheme based on protective agents for microgrid with parameters impact analysis. Scientific Reports, 15:44316, DOI:10.1038/s41598-025-28178-0.
Eissa MA, Samy MM, Said M (2026) Techno-economic analysis and resilience enhancement of a hospital microgrid under grid outage scenarios. Scientific Reports, 16:18358, DOI:10.1038/s41598-026-56638-8.
Nazartalab P, Alavi-Rad H (2026) Resilience-oriented optimization of hospital microgrids with critical load support using ESS and PV under grid outage conditions. Scientific Reports, 16:5475, DOI:10.1038/s41598-026-34992-x.
Hubner RU, Kusel C, Oestmann JW (2023) Resilience of the critical infrastructure in hospitals: categorization and quantification as a basis for optimization. Die Anaesthesiologie, 72(10):710-718, DOI:10.1007/s00101-023-01318-9.
Khalil M, Ravaghi H, Samhouri D, Abo J, Ali A, Sakr H, Camacho A (2022) What is "hospital resilience"? A scoping review on conceptualization, operationalization, and evaluation. Frontiers in Public Health, 10:1009400, DOI:10.3389/fpubh.2022.1009400.
