Passive Islanding Detection in Photovoltaic Based DC microgrid using Decision Parameters with Binary Tree

DC microgrid networks have prominence value due to the increase of power applications. However, these systems face protection challenges in the islanding scenario because most of the parameters such as reactive power and frequency are not present. Therefore, many techniques are not suitable like AC microgrids. This paper introducing a new passive islanding detection for photovoltaic (PV) based DC microgrid. During grid uncertainties, there arise overlappings in the system parameters because this detection is complex by single decision parameter and also the possibility of false outcomes. To avoid false detection results multiple decision parameters-based binary tree classifier is proposed in this paper. The decision parameters such as DC derivative (dIdc/dt), Teager energy operator (TEO), and harmonic amplification factor (HAF) are considered in this paper due to appreciable changes in islanding and other grid contingencies. The effectiveness of the proposed method is evaluated by MATLAB/Simulink platform. The effective islanding detection is ensured by considering the islanding scenario with other contingencies (load variations in DC and AC side and pole to pole (PP) fault occurring in DC side) in a result study where no major variations under normal operating conditions.