Disturbance Injection Based Decentralized Identification of Accidental Islanding

Unplanned islanding is intended to be identified due to the adverse effects on personnel and electrical equipment. In this paper, an active islanding identification technique is proposed based on disturbance injection through q-axis controller and an investigation technique, which incorporates measurement of mean of absolute frequency variation of the voltage signal passed through bandpass filter. The proposed technique is proven to be robust, offering decentralized islanding identification and being applicable to dynamic loading scenario, identifies island formation with zero nondetection zone for all the loads of quality factor (Q
_f
) ≤ 2.5, and does not destabilize the system on postislanding. The islanding identification time is <; 180 ms for all the loads of Q
_f
≤ 2.5, which is applicable to various international standards. The MATLAB2017b/Simulink platform is utilized to validate the efficacy of the proposed methodology.