Adaptive Kalman filter and neural network based high impedance fault detection in power distribution networks

Abstract

This paper presents an intelligent approach for high impedance fault (HIF) detection in power distribution feeders using combined Adaptive Extended Kalman Filter (AEKF) and probabilistic neural network (PNN). The AEKF is used to estimate the different harmonic components in HIF and NF (no-fault) current signals accurately under non-linear loading condition. The estimated harmonic components are used as features to train and test PNN for accurate classification of HIF from NF. Also a performance comparison is made between the results from feed forward neural network (FNN) and PNN for the same features extracted using AEKF. Thus a qualitative comparison is made for HIF detection and classification using the above techniques with FNN and PNN, separately. The testing results in noisy environment ensure the robustness of the proposed technique for HIF detection in distribution network.