Abstract: An example arc-fault detection apparatus includes an electrical current measuring device connectable to an electrical circuit and configured to provide a measured signal representing an electrical current flowing in the electrical circuit. A computing device is configured to: acquire a first plurality from samples of the measured signal included into a first time interval and a second plurality of sample from the measured signal included into a second time interval, subsequent to the first time interval; perform a discrete derivative of the first plurality of samples obtaining past derivative samples and a discrete derivative of the second plurality of samples obtaining present derivative samples; compute a derivative similarity index by processing the past derivative samples and the present derivative samples according to an auto-correlation type function quantifying the similarity of the past derivative samples with the present derivative samples.

Abstract: An arc fault detection circuit includes a current sensing circuit coupled to a line conductor carrying a current. The current sensing circuit operates to sense current and output data indicative of the sensed current. A processing circuit implements a frequency transform algorithm to transform the output data to frequency data in a low frequency range and with a high spectral resolution where a minimum short time observation window is concerned. The processing circuit identifies an arc fault condition on the line conductor by identifying differences in said frequency data between at least two subsequent observation windows and identifying characteristics which exceed thresholds.

Abstract: An arc fault detection circuit includes a current sensing circuit coupled to a line conductor carrying a current. The current sensing circuit operates to sense current and output data indicative of the sensed current. A processing circuit implements a frequency transform algorithm to transform the output data to frequency data in a low frequency range and with a high spectral resolution where a minimum short time observation window is concerned. The processing circuit identifies an arc fault condition on the line conductor by identifying differences in said frequency data between at least two subsequent observation windows and identifying characteristics which exceed thresholds.