Accurate Fault Classification in Series Compensated Multi-Terminal Extra High Voltage Transmission Line using Probabilistic Neural Network

This paper presents a novel method proposed for protection of multi-bus EHV transmission system which is employed with a series capacitor bank. It has been shown that for a wide range of operating conditions and fault scenarios arising in a series compensated transmission line poses great difficulty with conventional relaying scheme to correctly identify and classify faults. A Multi-resolution wavelet transform based approach is used to decompose the signals derived from CT in to low and high frequency components accompanied by a Neural Network scheme to classify the faults. The Probabilistic Neural Network (PNN) based scheme associated with a novel feature extraction methodology has been shown to classify the faults on the transmission line. Detailed and extensive simulation studies illustrates that the fault pattern classification approach presented here is effective and robust to handle wide range of operating conditions of Transmission line to which may undergo variety of fault conditions.