Electricity theft detection in low voltage networks with smart meters using state estimation

Electricity theft is a critical issue that needs to be solved. In the future smart meters can immediately monitor the entire system and obtain real-time operating data of a low voltage distribution system. By properly using the data from smart meters, the real-time distribution system model can be established and correctness of the customer electricity data can be guaranteed. This paper proposes a methodology based on distribution state estimation to detect customer tampered data. In order to obtain better results, a semi-definite programming method is used to solve this problem. The convex semi-definite relaxation (SDR) technique renders the non-convex robust state estimation (R-SE) problem be effectively solvable. Compared with the conventional Newton Raphson method, semi-definite programming can achieve the global optimal solution, rather than local optimal solutions. A low voltage distribution system is selected to validate the effectiveness of the proposed method. Test results indicate that the proposed method, which involves electricity theft detection, suspect confirmation, and data correction, can solve the non-technical loss problem, thus benefitting smart grid operations.