A Comprehensive Approach to Analysis and Operation Management of 0.4 kV Electrical Networks Based on Data from an Intelligent Electricity Metering System

Abstract

Enhancing observability and energy efficiency of 0.4 kV distribution networks is a priority objective in the digitalization of the power sector. Traditional solutions based on digital substations are economically unfeasible for numerous low-voltage facilities. This paper proposes a comprehensive approach utilizing data from the «KUMIRRESOURCE» ICT platform with a 1-minute sampling interval to address operational management tasks without installing additional hardware. Two interrelated tasks are addressed: multi-criteria optimization of single-phase consumer distribution across phases to minimize losses and asymmetry, and detection of anomalous consumption patterns (unmetered consumption, hidden cryptocurrency mining) using machine learning methods. A genetic algorithm is employed for phase load optimization, while time series analysis with adaptive threshold correction is used for anomaly detection. Simulation based on real data demonstrated a 21.2 % reduction in technical losses, complete elimination of current imbalance, and resolution of phase overloads. The results confirm the potential to transform the commercial metering system into a tool for active network operation management.