Angle-compensated non-invasive current measurement of multi-conductor cables via a TMR sensor-array

Real-time current monitoring of currents in smart grid poses significant challenges, particularly within multi-conductor cables where inherent phase imbalance and complex electromagnetic coupling distort the magnetic field patterns, complicating accurate measurement. Conventional non-invasive techniques are often applied to a single conductor, insufficient adaptability to multiple varying conductor arrangements, and require operational interruption for installation. To address these issues, this study designs an angle-compensated non-invasive current measurement method for multi-conductor cables with an annular TMR sensor array. A current inversion model incorporating conductor position and inclination parameters is established. A dual-stage optimization strategy combining Particle Swarm Optimization and the Levenberg–Marquardt Algorithm is proposed to solve the inversion problem efficiently. Experimental results demonstrate that the maximum current error remains below 7% under balanced three-phase conditions. The method also maintains high accuracy in the presence of conductor eccentricity, enabling reliable non-invasive current measurement across diverse multi-conductor configurations.