A Hybrid Islanding Detection Method Based on Pearson Correlation Coefficient

Abstract

Renewable Energy Resources (RERs) have become essential components in modern electrical power systems. However, one of the significant problems in grid-connected microgrid is unintentional islanding, which can result into outages on the main grid power supply in cases of faults or other uncertainties in the main network. This problem poses a danger to maintenance workers can cause damage to customers and microgrid equipment. Therefore, effective islanding detection methods are necessary to improve the operation of microgrids mitigate and tackle these effects. Previously hybrid proposed islanding detection methods such as the modified reactive power control have only focused on inverter-based sources. This paper proposes a hybrid technique for islanding detection with zero non-detection zone and low impact on power quality. A novel voltage fluctuation injection method based on Pearson correlation coefficient is proposed for islanding detection that involves monitoring the rate of change of Point of Common Coupling (PCC) voltage and switching a high impedance. The Pearson correlation coefficient between the three-phase PCC voltage and the three-phase voltage of switched impedance is used as the Islanding Detection (ID) index. The proposed method is implemented in a modified IEEE 13-bus test system using MATLAB Simulink. The normally operated grid-connected microgrid had the highest Pearson correlation coefficient. In contrast, the islanding operation had the lowest Pearson correlation coefficient. The method can thus successfully differentiate between islanding and non-islanding events, and islanding can be detected in less than 0.20 s. The proposed method is compared with the islanding detection method based-modified Reactive Power Control (RPC). The proposed method can differentiate between islanding and non-islanding successfully. In contrast, some non-islanding events are incorrectly detected by the RPC-based method.