MACHINE INTELLIGENT DIRECTIONAL PROTECTION OF TRANSMISSION LINE BASED ON ENSEMBLE DECISION TREE

Abstract

The proposed work establishes a distance protection scheme for a high voltage power transmission line, Using Principal Component Analysis (PCA) and Ensemble Decision Tree (EDT) methodologies. Despite major changes in power constrains and sophisticated design, distinguishing between internal, external fault, and power swing disorder is a highly stimulating effort in transmission line distance protection setup. As a foundation, this study used directional evaluations to build refined permissible reach in place of impedance, current amplitude, or voltage created approach. As a matter of fact, rather than the impedance or amplitude of current or voltage created method, this paper cast-off directional evaluation founded improved permissive overreach transfer trip (POTT) constructed method that employs waveform of local end three phase voltages and currents and remote ends tripping signal that integrates nominal bandwidth communication channel, and then PCA is cast-off to decrease. The dimensions of these voltages and currents samples are used to train and test various EDTs for recognition of internal, external faults, and power swings, and it also delivers EDT created backup protection in the event of a communication disaster that uses only local end voltages and currents for protection. To ensure relaying under various operational conditions, the power transmission line is visualized in PSCAD/EMTDC software. Throughout Power Transfer, the expected process is evaluated in MATLAB under various operational conditions, and the validated data is mirrored in PSCAD/EMTDC by manipulating the inception angle, fault location, fault resistances, power angle, and fault type all are significant factors. The results reveal that the proposed technique is minimal, persistent, discriminating, and reckless beneath unlike high power electrical transmission line operating conditions.