Fault Ascertaining Strategy for a Fixed Capacitor Compensated Transmission Network

Abstract

This paper presents an intelligent computing-based scheme for identifying the fault events in fixed capacitor compensated power (FCCP) network. The significant pattern change in the 3-phase current signal during shunt fault incident in the network has been utilized for identifying the kinds of particular events. The fault current samples are processed by discrete wavelet transform for acquiring the vital fault features. Subsequently, the elected feature vectors have been applied as input to the designed events classifier models (based on machine learning) for identifying the particular fault event. In the present work, both parametric and non-parametric machines learning techniques (quadratic discriminant analysis and weighted-K nearest neighbor) based classifier models have been formulated for fault events classification. The competency of the proposed integrated fault events ascertaining scheme has been evaluated for different circumstances on a simulated FCCP lines. The test results reasserted the strength of the proposed scheme in recognizing the individual shunt fault in the compensated power lines. It is seen that, the W-KNN based approach endows improved average accuracy level (99.55%) during ascertaining the events in FCCP network as compared previously reported methodologies. © (2024), (Greater Mekong Subregion Academic and Research Network). All Rights Reserved.