Detection Method for Current Transformer Metering Error of High-Voltage Line Group Based on Hierarchical Progression

The error stability of current transformers (CTs) during online operation is highly sensitive to power grid fluctuations, often resulting in error overruns that compromise the accuracy of power metering. To address the limitations of traditional offline CT calibration methods, this paper proposes a hierarchical, progressive detection framework for online metering error identification in group CTs, grounded in Kirchhoff’s Current Law (KCL) constraints. The approach begins by constructing multi-branch current vectors and formulating equations based on network topology to establish a CT error detection model. Robust Principal Component Analysis (RPCA) is then employed to extract abnormal features from CT measurement data, enabling rapid identification of out-of-range CTs. Finally, an optimization objective is formulated to minimize the absolute values of node current vectors under KCL constraints, and the Quantum Particle Swarm Optimization (QPSO) algorithm is applied to quantitatively estimate the amplitude and phase errors of group CTs. Unlike conventional outage-based calibration methods, the proposed framework incorporates physical constraints present during actual grid operation, enabling real-time detection and quantification of CT metering errors with 0.2-level accuracy.It helps maintain the fairness of electricity trade settlement, the reliability of relay protection actions, and the accuracy of system status estimation.