The Power Transformer Differential Protection Using Decision Tree

Abstract

This paper presents a new power transformer protection method by intelligent differential relay. This method is based on differential current sampling caused by primary and secondary three phase current of transformer. Detecting the type of fault in protection place is based on performance of the learned decision tree. The simulations of power system and differential relay are accomplished in PSCAD and MATLAB. The main characteristic of the relay is being intelligence such that at first step the decision tree is learned by several samples including faulted and inrush currents and then the relay detects the internal fault intelligently and quickly and the outputs of the relay order to the breakers in two ends of the transformer to be opened.

An intelligent differential protection of power transformer based on artificial neural network

“…Figure 3 illustrates relationship between the differential current and the restring current (operation relay characteristics) [17]. When the ratio of the pickup is bigger than the bias setting therefor, this ratio value will fall in the tripping region (positive region), otherwise if this ratio is smaller than the bias setting then this ratio value located fall in the blocking region (negative reign) [18], [19]. When faults occur outside of zone, the effect of restraining coils on the operating coils in the types of relays produces conflicting torque, preventing the relay from operating.…”