Research on a Novel Fault Identification and Rapid Location Method for Low-Voltage Distribution Lines

Fault diagnosis on low-voltage distribution lines is often limited by inaccurate classification, delayed localization, and sensitivity to interference. This study develops a fault-identification and rapid-location method that combines multisource feature analysis with Internet of Things (IoT) sensing. Short-circuit, grounding, and open-circuit faults are first characterized using line measurements. A wavelet-transform and random-forest model then classifies the fault type, while an improved weighted-centroid algorithm uses data from deployed IoT sensor nodes to estimate its position. The method was evaluated in simulations and on an operating line. Across these tests, fault-identification accuracy exceeded 95%, location error remained within 5m, and response time was below 2s. These results satisfy the stated engineering requirements for fault handling on low-voltage distribution lines.