Online Error Detection for Charging Piles Based on Extended Kalman Filter

When electric vehicles (EVs) are charged through charging piles, discrepancies may arise between the energy measured by the pile and the energy actually consumed, impacting the accuracy of billing. Therefore, periodic calibration of the energy metering performance of charging piles is crucial. Traditional on-site calibration methods based on the field calibrator and load devices suffer from low efficiency and high costs, while existing error calculation methods based on the energy conservation equations in the charging station struggle to effectively reflect the dynamic variations in error coefficients. To address these issues, an online error detection method for charging piles is proposed, utilizing the Extended Kalman Filter (EKF) to linearize nonlinear models locally and iteratively update error coefficients. Firstly, the topology of the charging station and the energy conservation equations are established. Next, building on the fundamentals of the EKF, the error characteristics are analyzed through the energy data of each charging pile. Finally, the effectiveness of proposed method is verified by case studies.