Comparative Analysis of Islanding Detection in Grid connected Distributed Generation with Variants of Decision Tree Approach

Islanding detection stands as a critical facet within grid-connected distributed generation (DG) systems. These systems harness the source of renewable energies such as solar photovoltaic (PV) panels, wind turbines, or small-scale generators, all interconnected with the primary electrical grid. The concept of islanding arises when a segment of the distributed generation system persists in supplying power to local loads, even when the primary grid has encountered a blackout or fault condition, resulting in its disconnection. Such a scenario can pose safety risks to utility personnel attempting grid repairs and can potentially cause equipment damage if not appropriately managed. This study undertakes a comparative analysis employing various decision tree approaches, namely the Coarse, Medium, and Fine Decision Tree Classifiers. Through extensive simulations, the efficacy of the proposed classifier is rigorously examined, and a comprehensive comparison is established.