Transformer differential protection using principal component analysis

This paper describes a new algorithm for transformer differential protection, based on pattern recognition of the differential current obtained as the phasor sum of the current transformers secondary currents. The algorithm uses principal component analysis (PCA) to pre-processing data from the power system, in order to eliminate redundant information and enhance hidden pattern in differential current to discriminate between internal faults from inrush and overexcitation conditions. The algorithm was proven using PSCAD/EMTDC simulations in a three-phase power system considering critical fault cases. The results show the feasibility to implement this algorithm for transformer differential protection