Detection of Series DC Arc on a Distribution Node using Discrete-Time Parameter Identification Techniques

Discrete-time parameter identification techniques are studied extensively and are implemented to detect and localize the series arc discharge on a multi-load distribution node. The techniques include Recursive Least Squares (RLS) and the Kalman Filter (KF) which by estimating the line parameters - resistance and inductance will help monitor their condition. The systems electrical components vary abruptly during an arc discharge on a line of the grid. This disturbance from the arc discharge also travels to the adjacent loads and lines. The RLS and KF will then act as detectors functioning on a line and consequently help localize the faulted line/load. The two mentioned techniques delineate the variation in values of the grid parameters even for a large noisy arc.