Electricity theft detection using neural networks and Particle Swarm Optimization optimization

Electricity theft poses significant challenges to the integrity and sustainability of power systems. In this study, we address this issue by combining neural networks with Particle Swarm Optimization (PSO) algorithm for effective detection of electricity theft across different risk levels. Using three datasets representing high, medium, and low-risk theft users, we trained and optimized a neural network to achieve rapid convergence and excellent predictive accuracy, as demonstrated by iterative and accuracy curves. Our results showcase the robustness of the PSO-optimized neural network in accurately detecting theft phenomena regardless of user density, highlighting its applicability in diverse scenarios. The study’s findings not only validate the effectiveness of our approach but also pave the way for future enhancements in electricity theft detection, such as optimizing parameters and incorporating deep learning techniques. Overall, this research contributes to advancing the field of electricity theft detection, ensuring the safety and reliability of power grids.