Initial Features of the Unintended Atmospheric Pressure DC Arcs and Their Application on the Fault Detection

Abstract

Unintended atmospheric pressure dc arc is very common and harmful in many power applications. There is still no effective detection method for series dc arcs. In this paper, the initial features of current and electromagnetic signal of series dc arc are investigated on an arc generation platform. It is found that the current varies abruptly at the starting moment of the arc, and then a series of stochastic pulses can be observed. Corresponding to the current pulses, electromagnetic pulses are also measured by antenna simultaneously. Based on the abrupt current variation and unique pulse-time pattern features of the series dc arc, a fault detection algorithm is proposed. The proposed algorithm is proved to be robust and effective.