Controlled islanding based on slow-coherency and KWP theory

In power system emergency control, the controlled islanding is the last measure when it faces the severe cascading contingencies. Its objective is to maintain the stability of the sub partitions and reduce the total loss of load of the whole system to the minimum. A dynamic partitioning and controlled islanding strategy in power system emergency control has been investigated in this paper. Firstly, according to slow coherency grouping algorithm, generator with similar oscillation mode will separate into different optimal groups, and aggregate these generator groups and corresponding buses into some virtual coherent nodes. On the basis of the Laplace partitioning strategy, a novel weight index has been proposed to measure the “distance” of two adjacent nodes and improve the Laplace partitioning strategy. Then a novel heuristic neighborhood searching partitioning strategy has been raised to search the possible cutsets with the smaller net generation-load imbalance. Finally, a heuristic load-shedding approach has been applied on all these possible cut-sets to calculate the amount of load-shedding of each partition. The applications on 39-bus New England test system and the IEEE118 system validate the effectiveness of the improvement of controlled islanding strategy based on slow coherency and KWP theory proposed in this paper.