Polar features analysis of electromagnetic field signatures for non-intrusive transmission line fault diagnosis
Year: 2026
Type: article
Abstract: Conventional power system protection schemes rely on contact-based voltage and current measurements, which are often constrained by current transformer saturation, insulation requirements, and installation complexity. This paper presents a non-intrusive fault diagnostic framework based on electromagnetic field signatures derived from transmission line currents. A current-to-magnetic-field transformation is used to obtain non-contact magnetic field components, which are represented in the polar domain using radial magnitude (ρ) and angular orientation (θ). The polar representation enables intuitive visualization of fault-induced electromagnetic disturbances and supports systematic feature extraction. The electromagnetic validation of cross-platform is accomplished by matching the results of analytical MATLAB/Simulink analysis to finite element analysis in COMSOL Multiphysics, showing consistency in the magnitude and spatial behavior of the magnetic field. Various such fault studies are carried out at different points along the transmission line. Statistical analysis indicates that radial component features obtained as features of the radial component, distinguish well between faults and attenuate predictably with fault distance, making them distance-sensitive diagnostics. Angular features on the contrary are not quantitatively sensitive enough and are thus used in qualitative interpretation. The findings validate the hypothesis that polar-domain magnetic field sensors based on non-contact sensing are a physically realistic, dependable, and scalable method of diagnosing transmission line faults and that the family has a significant prospect of intelligent protection in the future.
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7127048734