A decision tree based approach for controlled islanding of microgrids

This paper presents a decision tree based systematic approach for controlled islanding of grid-connected microgrids. The objective of the proposed approach is to develop an adaptive controlled islanding methodology to be implemented as a preventive control component in emergency control strategy for microgrid operations. A contingency-oriented decision tree classifier is trained with event database generated from offline simulations of system events. The trained decision tree classifier is capable of identifying system events that warrant controlled islanding of microgrids as a preventive control measure. Real time voltage and current measurements are utilized in conjunction with the trained decision tree classifier for online decision support on controlled islanding strategy of microgrids. A microgrid test system model consisting of multiple distributed generations and energy storage systems is employed to demonstrate the performance of the proposed approach.