The article provides a description of the microgrid function optimization model for solving the problem of reducing the cost of electricity purchases by the owner of such a microgrid on the retail electricity market. The objective function of minimizing the cost of electricity purchase has been formalized. Approaches to determining the coefficients of penalty functions designed to cut off technically unacceptable modes of the microgrid have been determined. A mathematical model of an energy storage system (ESS) is proposed, in which the functions of calculating the technological costs of electricity are formalized. A system of constraints for the problem of minimization of the cost of electricity purchase is given. Approaches to solving the problem of evaluating the effect of SPP and ESS operation in microgrids using optimization results are determined. Examples of microgrid operation optimization for an enterprise with its own SPP and ESS are presented. The need for additional control of the economic expediency of involving ESS in the regulation of electricity balances at the enterprise was noted. References 17, figures 3, table 1.

This research presents an advanced methodology for smart management of energy losses in electrical distribution networks by leveraging deep neural network architectures. The primary objective is to enhance the accuracy of short-term forecasting for nodal loads and corresponding energy losses, enabling more efficient and intelligent grid operation. Two predictive approaches were explored: the first involves separate forecasting of nodal loads followed by loss calculations, while the second directly estimates network-wide energy losses. For model implementation, Long Short-Term Memory (LSTM) networks and the enhanced Residual Network (eResNet) architecture, developed at the Institute of Electrodynamics of the National Academy of Sciences of Ukraine, were utilized. The models were validated using retrospective data from a Ukrainian Distribution System Operator (DSO) covering the period from 2017 to 2019 with 30 min sampling intervals. An adapted CIGRE benchmark medium-voltage network was employed to simulate real-world conditions. Given the presence of anomalies and missing values in the operational data, a two-stage preprocessing algorithm incorporating DBSCAN clustering was applied for data cleansing and imputation. The results indicate a Mean Absolute Percentage Error (MAPE) of just 3.29% for nodal load forecasts, which significantly outperforms conventional methods. These findings affirm the feasibility of integrating such models into Smart Grid infrastructures to improve decision-making, minimize operational losses, and reduce the costs associated with energy loss compensation. This study provides a practical framework for data-driven energy loss management, emphasizing the growing role of artificial intelligence in modern power systems.

The main differences in pricing and tariffing for industrial consumers of electricity with different forms of electricity metering are considered. Based on the analysis of tariff formation for the final consumer of electricity, components are identified that have a significant impact on the results of solving the problem of assessing the feasibility of the consumer's transition to hourly electricity metering. Such components include the cost of purchasing electricity in the market segment "day ahead" and the cost of accrued imbalances. The relative daily profile of electricity consumption is considered in order to study the influence of the features of the daily load schedule on the weighted average daily market price of electricity. The importance of estimating the cost of daily load profiles when comparing the cost of electricity for the consumer in the group with integrated electricity metering and in terms of individual hourly metering is substantiated. The effect of underestimation of volumes and value of imbalances in the group with integrated electricity metering in comparison with hourly accruals of volumes and value of imbalances is theoretically substantiated. The main components for comparative assessment of the expediency of the consumer's exit from the group with integrated metering of electricity and the transition to its hourly metering according to the individual daily load schedule are identified. Mathematical models for comparative calculations are developed. The use of these models allows to make an economically justified decision on the expediency of the consumer leaving the group without hourly metering of electricity to the model of purchasing electricity with hourly metering. The main approaches to such an assessment are demonstrated on the example of calculations for an industrial enterprise in some regions of Ukraine. Bibl. 15, fig. 3.

Electricity theft is a critical source of non-technical losses in modern power systems, causing substantial financial and operational challenges for utilities. Traditional detection methods, such as manual inspections, are inadequate to detect advanced theft techniques, including meter tampering and cyberattacks on smart grids. This study introduces a machine learning-based framework for electricity theft detection using the TDD2022 dataset (derived from OEDI) and evaluates multiple algorithms—Random Forest, Decision Tree, XGBoost, LightGBM, CatBoost, Extra Trees, and Logistic Regression. To address class imbalance, SMOTE is applied, while feature selection leverages LASSO and ReliefF. Experiments compare electricity-only data with multi-utility inputs (electricity and gas) under balanced and imbalanced conditions. Results show that tree-based ensembles, particularly Extra Trees combined with SMOTE and ReliefF, achieve superior performance (accuracy $>95\%$, AUC $\approx 0.99$). Consumer-specific models outperform global models, with commercial classes yielding near-perfect detection, while residential profiles remain challenging. The findings highlight the importance of tailored modeling and feature selection for scalable, accurate theft detection in smart grid environments.