A new passive islanding detection method based on rate of change of frequency

As the penetration of distributed generation (DG) continues to rise, distribution networks face an increased risk of multi-DG islanding effects. To safeguard power quality and electrical safety, inverters must incorporate islanding detection capabilities. To address the shortcomings of traditional passive islanding detection methods, which suffer from low accuracy and large detection blind zones, a novel islanding detection approach based on the rate of change of frequency is introduced. This method leverages a frequency-locked loop to control reactive power, enabling grid frequency tracking through reactive power negative feedback. Initially, a high-pass filter is utilized to isolate the high-frequency components of the frequency. Subsequently, a locally weighted linear regression algorithm is employed to fit these high-frequency components of the frequency change curve, and the slope of the fitted curve is computed. Ultimately, islanding detection is accomplished by analyzing the fluctuations in the slope of the frequency curve at the point of common coupling before and after islanding occurs. MATLAB/Simulink simulation results validate that this method can precisely detect islanding effects across diverse scenarios, eliminating detection blind zones and demonstrating robust applicability.