Auto‐correlation‐based Islanding detection technique verified through hardware‐in‐loop testing

Abstract

This study describes a novel Islanding detection technique based on a discrimination factor (DF) which is derived from an auto-correlation factor (ACF) of the voltage signals acquired from the terminal of the distributed generator. The value of DF is calculated from the most affected lags of the ACF during various Islanding and non-Islanding conditions. Further, the validation of the presented scheme has been carried out by developing a hardware-in-loop laboratory setup. In this setup, a power distribution network has been modelled on a real-time digital simulator (RTDS/RSCAD) which is physically connected to the digital signal processor controller. Various non-Islanding and Islanding test scenarios have been generated. The proposed ACF-based technique is able to differentiate Islanding state with non-Islanding states even for perfect power equilibrium situation. The proposed approach can determine the Islanding state in a period of three cycles from the inception of Islanding. Finally, a relative assessment of the above algorithm with other existing methods shows its supremacy in terms of better stability during critical non-Islanding situations and lower non-detection zone in case of Islanding states.