Electrical Signature Analysis for Open Circuit Faults Detection of Inverter With Various Disturbance

This article proposes an electrical signature analysis method for open-circuit faults (OCFs) detection of inverter with various disturbances in distribution grid. According to the fault mechanism, the fundamental value, rated harmonics, and direct current component of three-phase currents are used as fault electrical signatures. The signatures are estimated by unscented Kalman filter (UKF) and recognized by extreme learning machine (ELM) for fault detection. Both OCF of single switch and OCFs of multiple switches are tested with consideration of direct disturbances such as load change, overcurrent and bias current, and indirect disturbances such as grid frequency variation, background harmonics, and unbalanced voltage dip. The simulations and experiments show that the OCF detection of the new method is still accurate even with these disturbances, and reveal that only the signatures of the faulty phase current are immune to the disturbances while the ones of unfaulty phases are not. The robustness when facing the various disturbances and all explainable detection results make the new method suitable and effective for OCF of inverter detection in complicated distribution grid environment.