RFMFA Technique for Islanding Detection Scheme in Integrated DG System

Abstract

EEE 1547 specifies time limits for the completion of DG interconnection tasks, and one of these tasks is the detection of islanding, which must be completed within two seconds of the task's start. This thesis provides new and effective islanding detection algorithms for a hybrid distributed generating system connected to the grid. The suggested technique is a hybrid of two different optimization algorithms: Random Forest (RF) and Moth Flame Optimization (MFO). At the objective stage, this technique employs the ROCOF. The suggested methods are distinguished by their ease of use due to the lack of a mandatory minimum. Furthermore, they are unaffected by the capacity or kind of Distributed Generation linked to the utility grid and not affected by system factors. In addition, the NDZ is not present after applying these processes. When compared to the standard approaches, these methods show a marked improvement in their ability to classify events into islanding and non-islanding categories. Several elements, including varying load situations, switching operations, and network conditions, must be considered before determining the proposed solutions' viability. The MATLAB/Simulink platform provides a setting where the suggested procedures may be implemented.