Kalman Filter for Detecting Serial Arc Faults in a Domestic Electrical Network

The detection of arc faults in domestic networks has been an important subject for industrials and researchers for many years. The challenge for fault detection algorithms is to work efficiently with different circuit configurations where serial arc faults are difficult to identify. In this work, we focus on arc fault situations in the presence of masking loads, while taking into account the transient effect of starting loads. The proposed algorithm is based on a Kalman filter. Both the estimated current at the Kalman filter output and its state X2 are used to generate fault symptoms. Then a detection logic block based on fuzzy logic block confirms the presence of the arcing fault. The algorithm is tested on appliances in single or masked load configurations selected according to the requirements of UL 1699 and IEC 62606. The algorithm is also tested in steady state or at load start-up (transient state). The performance of this method is also studied and discussed. Experimental results show that the method we propose can effectively detect arcing faults, avoiding false tripping.