Robust modified passive islanding detection for microgrids using mathematical morphology based dual algorithm

The integration of distributed generation (DG) in microgrids has brought the challenge of islanding detection, where a portion of the grid operates independently due to disconnection from the main utility. Traditional islanding detection methods often struggle to balance detection speed, reliability, and non-detection zones (NDZ). This paper presents a novel modified passive islanding detection strategy based on a mathematical morphological filter (MMF) with a sliding window method-based median filter (SWMBMF), designed to address these challenges in microgrid systems. The measured noisy voltage signal at point of common coupling (PCC)/DGs-terminal is initially estimated via SWMBMF by continuously monitoring the signals with minimal latency. Then, the MMF is utilized to process the estimated voltage to compute the voltage residuals index (VRI). Moreover, the VRI are compared with pre-specified threshold setting to detect islanding conditions, VRI also effectively distinguishes between normal and islanding conditions. Comprehensive MATLAB/Simulink 2023b simulations demonstrate the robustness of the proposed strategy under various islanding scenarios and grid disturbances, proving its effectiveness in ensuring the stability and reliability of microgrid operations. The proposed method demonstrates an impressive accuracy of 99%, successfully identifying islanding events within 5 ms. This rapid detection enhances grid reliability and minimizes risks associated with delayed islanding detection. The proposed scheme has negligible non detection zone (NDZ).