A K-Nearest Neighbour based Protection Scheme for UPFC Compensated Double Circuit Transmission Line with Wind farms

Abstract

This paper proposes a K-nearest neighbour (KNN) based protection approach for fault detection and classification for a shunt compensated three-phase transmission line integrated with wind-farms. Quite often, classical relays maloperate while protecting the transmission line which has been compensated with FACTs devices like unified power flow controller (UPFC). The protection of the transmission line is further complicated when it is integrated with the wind farm. The scheme initiates with the processing of raw time domain signals acquired from the relaying buses using discrete Fourier transform (DFT) for extracting fundamental component of voltage and current signals. For training of KNN, the fundamental component of current and voltage signal at different fault scenarios with the varying fault parameters are used as input. The effectiveness of the proposed method is tested for different fault scenarios involving variation in the fault type, fault resistance, fault location, and fault inception angle. The test results confirm the suitability of the proposed scheme for UPFC compensated transmission line with wind farm integration.