Robust Hierarchical Traveling-Wave Partial Discharge Localization for Efficient Insulation Defect Detection of Medium-Voltage Overhead Lines

Medium-voltage overhead lines typically span kilometers, yet classical partial discharge (PD) detection methods are limited to meter-scale ranges, resulting in inefficient inspection. Traveling-wave PD localization provides a kilometer-scale monitoring solution that addresses this fundamental mismatch in spatial scale. However, its field deployment is hindered by insufficient detector sensitivity and severe reliability degradation caused by noise, frequency-dependent attenuation, and wave reflections in complex line environments. To address these challenges, this paper introduces a “global-to-local” detection methodology that enables both large-scale screening and equipment-level identification through hierarchical time-difference-of-arrival and polarity analysis. A dedicated detector has been developed, incorporating a non-invasive high-frequency current transformer and an ultra-low-noise amplifier capable of resolving sub-100 μA PD pulses, all synchronized via the Global Navigation Satellite System for multi-terminal acquisition. A dual-domain processing framework further enhances system robustness by integrating autocorrelation-guided reflection suppression in the time domain with adaptive subband fusion in the frequency domain, compensating for frequency-dependent attenuation while suppressing interference. Validation through simulation, laboratory tests, and field deployment confirms accurate defect localization under realistic noise and reflection conditions, demonstrating the practical viability of the system.