Performance assessment of IMF-based passive islanding detection method in various islanding and non-islanding events

This paper evaluates the performance of a passive islanding detection method in various islanding and non-islanding events and different network conditions. The method is based on calculating the intrinsic mode functions (IMFs) of the voltage of point of common coupling (PCC) using the empirical mode decomposition (EMD) technique. The performance of the method is investigated in two sample power systems in various conditions, such as UL 1741 test cases, existence of noise, unbalanced and nonlinear load, variations of load quality factor and load resonant frequency, presence of inverter-based and non-inverter-based DGs, induction motor starting and load switching event. The simulation results show that the method can effectively detect islanding in all cases. Moreover, it can properly distinguish between the islanding condition and non-islanding events.