Adaptive detection of DC Arc faults based on hurst exponents and current envelope

In this paper, an improved adaptive method for dc arc faults detection is proposed based on current envelope and Hurst exponents. The method is developed to be used in dc electrical distribution systems (EDS) including PV systems, DC micro-grids, and 48 V mild hybrid electric vehicles. The proposed method is shown to be effective in arc detection and immune to switching harmonics, high di/dt transients, sudden load changes, and noisy load profiles. A test setup is built and several hundred dc arcs were generated under different operation conditions including air gap, current, and various electrode thicknesses.