A Novel Adaptive Algorithm to Identify Inrush Using Mathematical Morphology

A novel scheme based on mathematical morphology to distinguish the inrush current from the internal fault current in power transformers is proposed, which is derived from high non-linearity in exciting cores of transformers. Transient current signals are extracted by use of the novel morphological gradient and morphological opening and closing transform, meanwhile various kinds of additive white noises are effectively suppressed by adaptive filtering. Based on different characteristics of transients between the short circuit current and the inrush current, the transformer criterion is developed. The proposed technique is immune to CT saturation. A total of 120 Electrical Power Dynamic Laboratory (EPDL) simulation cases are used to test the performance of the technique. This method has stability during the symmetrical inrush current and can be implemented in real time with very high accuracy. The results of different EPDL testing systems verify the feasibility and practicability of the proposed scheme