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.

This article proposes an approach for the reduction of electricity costs in residential microgrids using Factored Multi-Agent Centralised Policy Gradients (FACMAC). FACMAC is a multi-agent deep reinforcement learning algorithm, which employs a centralised but factored critic enabling more efficient scaling in tasks involving a large number of agents and/or high-dimensional state-action spaces. This makes FACMAC suitable for addressing the challenges associated with the complexity of microgrid optimisation. Additionally, this article introduces a stochastic model based on Probability Density Functions to address uncertainties in household power profiles, solar generation, and the operational parameters of electric vehicles and battery energy storage systems. The proposed approach also considers energy efficiency and cost reduction by incorporating the life cycle and installation costs of energy sources, including PV generation, alongside costs associated with electric vehicles and battery energy storage. A comprehensive evaluation validates the model's effectiveness across different scenarios. Simulation results demonstrate that the FACMAC algorithm can achieve up to 20% in cost reduction compared to the state-of-the-art multi-agent algorithm.

This article proposes an approach for the reduction of electricity costs in residential microgrids using Factored Multi-Agent Centralised Policy Gradients (FACMAC). FACMAC is a multi-agent deep reinforcement learning algorithm, which employs a centralised but factored critic enabling more efficient scaling in tasks involving a large number of agents and/or high-dimensional state-action spaces. This makes FACMAC suitable for addressing the challenges associated with the complexity of microgrid optimisation. Additionally, this article introduces a stochastic model based on Probability Density Functions to address uncertainties in household power profiles, solar generation, and the operational parameters of electric vehicles and battery energy storage systems. The proposed approachalso considers energy efficiency and cost reduction by incorporating the life cycle and installation costs of energy sources, including PV generation, alongside costs associated with electric vehicles and battery energy storage. A comprehensive evaluation validates the model’s effectiveness across different scenarios. Simulation results demonstrate that the FACMAC algorithm can achieve up to 20% in cost reduction compared to the state-of-the-art multi-agent algorithm

This article proposes an approach for the reduction of electricity costs in residential microgrids using Factored Multi-Agent Centralised Policy Gradients (FACMAC). FACMAC is a multi-agent deep reinforcement learning algorithm, which employs a centralised but factored critic enabling more efficient scaling in tasks involving a large number of agents and/or high-dimensional state-action spaces. This makes FACMAC suitable for addressing the challenges associated with the complexity of microgrid optimisation. Additionally, this article introduces a stochastic model based on Probability Density Functions to address uncertainties in household power profiles, solar generation, and the operational parameters of electric vehicles and battery energy storage systems. The proposed approach also considers energy efficiency and cost reduction by incorporating the life cycle and installation costs of energy sources, including PV generation, alongside costs associated with electric vehicles and battery energy storage. A comprehensive evaluation validates the model's effectiveness across different scenarios. Simulation results demonstrate that the FACMAC algorithm can achieve up to 20\% in cost reduction compared to the state-of-the-art multi-agent algorithm.