Ensuring the functional robustness of monitoring systems for autonomous energy complexed based on machine learning methods

DOI: 10.31673/2412-9070.2026.025711

Abstract

The article addresses an urgent scientific and practical problem of ensuring the functional resilience of monitoring systems for autonomous power complexes under conditions of dynamic technical failures, systemic anomalies, and critical loads. The relevance of the study is driven by the acute need to preserve the integrity of telemetry data during the unstable operation of power infrastructure, where classical routine backup methods demonstrate excessive inertia and fail to provide the necessary reaction speed to rapid changes in network parameters.
An intelligent software system has been developed that combines predictive diagnostics and adaptive real-time backup functions. The scientific novelty of the work lies in the synergistic use of two machine learning algorithms: Isolation Forest for instantaneous detection of current anomalies in multidimensional data arrays and Random Forest for forecasting future critical states based on the analysis of historical patterns. Such a comprehensive approach allows for the implementation of a proactive management strategy, initiating data preservation before the occurrence of an actual failure.
The software implementation of the system is based on a three-tier architecture and includes modules for intelligent analysis, automated backup planning, and a role-based access model. Experimental testing on real telemetry datasets confirmed the high efficiency of the system: the predictive analysis module ensures stable operation even under peak loads, which guarantees high functional resilience and data integrity in autonomous power networks.

Keywords: functional stability, autonomous power systems, software architecture, intelligent monitoring, machine learning, anomaly detection, predictive analytics, adaptive data backup.