Closed-Loop Health Management System of Relay Protection Device Based on Multi-Modal Perception and Dynamic Target Test
DOI:
https://doi.org/10.71222/94pmqp22Keywords:
relay protection device, health management system, multi-modal perception, predictive maintenance, condition monitoringAbstract
Conventional relay protection testing systems are plagued by a critical limitation: they prioritize functional compliance over long-term health assessment, leading to undetected mechanical wear, component aging, and failures caused by environmental stress. Routine maintenance based on time intervals, rather than actual device conditions, often results in over-maintenance or missed faults, and the lack of data-driven life prediction leads to inefficient asset utilization. To address these challenges, this paper proposes a closed-loop health management system for relay protection devices, introducing three core innovations: a quantitative health evaluation model based on multi-parameter fusion, a multi-modal perception layer that captures degradation signals in real time, and an intelligent decision-making mechanism at the edge. This system enables full-process monitoring, root-cause diagnosis, and adaptive maintenance, shifting the protection paradigm from passive fault response to proactive immunity, and laying the foundation for predictive maintenance and intelligent lifecycle management of protection assets.
References
1. J. Shen, M. Morrison, H. Miao, F. Gu, et al., "Harnessing deep learning for fault detection in Industry 4.0: A multimodal approach," in Proc. 2024 IEEE 6th Int. Conf. Cognit. Mach. Intell. (CogMI), 2024, doi: 10.1109/CogMI62246.2024.00045.
2. T. Zhu, Y. Ran, X. Zhou, Y. Wen, et al., "A survey on intelligent predictive maintenance (IPdM) in the era of fully connected intelligence," IEEE Commun. Surv. Tutor., vol. PP, no. 99, pp. 1–1, 2025, doi: 10.1109/COMST.2025.3567802.
3. S. D. Robinsha and B. Amutha, "Interfacing multi-modal AI with IoT: Unlocking new frontiers," in Multimodal Generative AI, Singapore: Springer Nature, 2025, pp. 323–346, doi: 10.1007/978-981-96-2355-6_14.
4. H. Dhungana, "Biomimicry as a decision-making methodology in condition monitoring," Front. Artif. Intell., vol. 8, 2025, Art. no. 1485489, doi: 10.3389/frai.2025.1485489.
5. D. Guidotti, L. Pandolfo, L. Pulina, et al., "A systematic literature review of supervised machine learning techniques for predictive maintenance in Industry 4.0," IEEE Access, vol. PP, no. 99, pp. 1–1, 2025, doi: 10.1109/ACCESS.2025.3578686.
6. C. M. Walker, L. Lin, V. Agarwal, N. J. Lybeck, A. Hall, et al., Demonstration and Evaluation of Explainable and Trustworthy Predictive Technology for Condition-based Maintenance, Idaho National Laboratory (INL), Idaho Falls, ID, USA, Tech. Rep. INL/RPT-24-80727-Rev000, 2024, doi: 10.2172/2474859.
7. W. Zhao, A. Cui, M. Fang, L. Yang, K. Chen, et al., "State assessment of 110–220 kV intelligent substation based on multi-sensor fusion algorithm control and image vision," Front. Energy Res., vol. 10, 2023, Art. no. 1047359, doi: 10.3389/fenrg.2022.1047359.
8. Y. Wang, G. Sun, Z. Sun, J. Wang, J. Li, C. Zhao, et al., "Toward realization of low-altitude economy networks: Core archi-tecture, integrated technologies, and future directions," arXiv preprint arXiv:2504.21583, 2025.
