Modal voltage decomposition-based passive method for islanding detection using variational mode decomposition in active distribution network

Abstract

Ascertaining unintentional islanding conditions is a challenging task for power utility operators due to the pervasive penetration of Distributed Energy Resources (DER). The presence of DER has a significant adverse impact on islanding detection schemes for various types of Distributed Generators (DG) and inverters. This paper addresses the passive islanding detection mechanism based on the mode energy index using Variational Mode Decomposition (VMD) in an Active Distribution Network (ADN). The prime attributes of VMD are noise immunity, robustness, and non-recursive signal processing, making it more promising and competent for signal decomposition than other extant signal processing methods. In this paper, three-phase instantaneous voltages at the target DG end have been acquired to compute modal voltage signals, and VMD has been utilized to decompose the obtained signals into four modes. Thereupon, the energy was calculated for the selected mode. In the context of threshold selection, a series of simulation is carried out for various islanding and non-islanding events. The energy threshold is then selected by implementing “Otsu thresholding method”. The estimated energy index has ultimately been utilized as an islanding detection indicator. In this work, two different test systems have been modelled and simulated using EMTDC/PSCAD. The simulation result outcomes asserted that the developed approach effectively ascertains the islanding and non-islanding events. Furthermore, the proposed islanding detection technique has been validated using the Real-Time Digital Simulator (RTDS) and dSPACE MicroLabBox based co-simulation platform. © 2023