Online Short-Circuit Fault Diagnosis in Three-Core Power Distribution Cable Based on Magnetic Pattern
Jingyi Liu, Chi-Kwan Lee, Philip W. T. Pong
Abstract
The short-circuit fault in three-core power distribution cables is the leading cause of failure in the power distribution network. It is imperative to diagnose it as soon as possible to avoid the consequent power outages. Against the common fault diagnosis methods that either necessitate the three-phase current or the disconnection of cable from the grid, this article proposes an online fault diagnosis method by measuring the tangential magnetic field around the power cable surface in a nonintrusive way. The formulated magnetic pattern is first introduced to detect and classify short-circuit faults. A normal operation results in a circular magnetic pattern, while each type of short-circuit fault uniquely changes the magnetic pattern. In general, short-circuit faults result in a remarkable change in the radius of the magnetic pattern, which can be detected within five samples based on the cumulative sum-based algorithm. The sensitivity of the proposed method on load variation, homogeneous components, and cable structure is also evaluated. Both simulation and experimental results validate the effectiveness and robustness of the proposed method.