Protection scheme for power transmission lines based on SVM and ANN considering the presence of non‐linear loads

Abstract

The proliferation of non-linear loads in recent times has a strong bearing on power quality due to the presence of harmonics in the power system. Harmonics lead to erroneous measurement and malfunctioning of protective relays, thus reducing the efficacy of transmission line protection. With the aim of improving the dependability of the protection scheme under varying non-linear loading condition, the study presents a hybrid support vector machine (SVM), artificial neural network (ANN) and Kalman filter based algorithm using voltage harmonics for the protection of three-phase transmission line. The post-fault voltage signals are processed by a Kalman filter to estimate the harmonic components, which serve as feature vectors for performing the fault detection and classification by SVM and zone identification as well as the location by ANN. The harmonic information discriminates faults from disturbances based on variations in the fundamental component, to improve the selectivity and accuracy. Considering the stochastic nature of fault occurrence in power systems, the efficacy of the proposed scheme is validated using Monte Carlo simulation. The experimental results confirm the superiority of using voltage harmonics for improving the dependability of transmission line protection.