Study Of The Micro-grid Islanding Detection Based On Improved Active Frequency Drift Algorithm

With the rapid growth of electricity load, the problems of energy shortage andenvironmental pollution are getting worse, more and more countries and regions begin tofocus on the development of micro-grid, micro-grid power system has become the mainstreamtrend of development. Unintentional islanding is one of the problems that can`t be ignored.The frequent islanding cause harm not only on equipment life, but also on the safety ofmaintenance personnel.This paper introduces the main problems of the micro-grid system and island testingstandard, islanding detection model and existing islanding detection method, and putemphasis on the analysis and research of proactive islanding detection on active frequencydrift (AFD) and active frequency drift with positive feedback (AFDPF). The traditionalmethod of AFD is that the non-detection zone will be decided because of the limit by the totalharmonic distortion (THD) when the applied disturbance is injected to the grid, and theislanding detection would be easily failed when frequency drift direction caused bydisturbance signal is opposite to the direction of load resonant frequency shift. Aiming at theshortcomings of the traditional method of initiative frequency offset, firstly, this paper designsa new current disturbance injection method to improve the disturbance waveform, Comparingwith the traditional method of AFD, the total generated harmonic distortion is reduced by30%, and the non-detection zone and detection rate are improve as well. Secondly, this paperapplies the improved disturbance waveform to the positive feedback active frequency driftmethod, and determines the direction in line with local load characteristics according to thespecific migration condition. This paper injects disturbance semaphore continuously and putforward with positive feedback active frequency drift method with the function of estimatingthe properties of load. This method avoids the influence of the load properties of islanddetection, reduces the total harmonic distortion of the system, and also quickly and efficiently.Finally, this paper builds a system modeling and makes a simulation for the proposedmethod by using Matlab/Simulink. The simulation results show that the total harmonicdistortion of the new method has been significantly reduced (nearly30%). When islandphenomenon occurs, the new method can quickly determine the load properties and applyconsistent disturbance signal, it also can detect island phenomenon rapidly, and thenon-detection is smaller.