A new hybrid intelligent method to avoid failure of power transformer differential protection relay

• The maloperation of the differential protection is investigated during CT saturation. • A new method is presented to improve the differential protection operation. • A new feature is suggested based on the analysis using the VMD transform. • The method prevents trip signals during transformer energizing with CT saturation. • The method detects internal faults even in the presence of CT saturation. The differential relay is used as the main protection for the power transformers with greater than 10 MW electrical power. However, this equipment is subject to operational failure under inrush current conditions or saturation of current transformers (CTs) during external faults. In this paper, a new method is presented based on the combination of signal analysis and artificial intelligence for power transformer protection. At first, the non-stationary differential current is analyzed by the variational mode decomposition (VMD) technique. Then, a new feature is extracted based on decomposed modes. This feature, along with bias and differential currents, is considered one of the three main attributes to detect internal faults from inrush currents and external faults. The extracted features from different operating conditions of the power transformer are used for training of different classifiers as the main detection core of the proposed protection method. The obtained results justify that the internal faults can be discriminated accurately from other conditions by the proposed method, even when CTs are deeply saturated. It should be noted that, in this study, A real 230/63 kV power transformer is simulated in the PSCAD environment, and different effective parameters on results are considered. On the other hand, the CT saturation is considered using the precise Jiels-Atherton model.