A Controlled Power System Splitting Algorithm Using PMU Data

The rapid growth of the load demand especially in industrial, residential, and commercial sectors pushes the power systems to their operational limits. Hence, the power system will be vulnerable to severe disturbances, which can be ended with undesired islanding or system blackout. Special Protection Schemes (SPS) are considered to prevent such incidents. Controlled islanding is an SPS that is considered as the last effort to prevent system blackout. In this paper, a new coherency method based on the fuzzy-c-means clustering algorithm is proposed. Besides, the Silhouettes algorithm is employed to find the optimal number of the coherent group. Additionally, the novel idea of island contraction is proposed. Center-Of-Inertia concept, fault detection, and detection of affected generators are utilized to discover whether the number of islands can be reduced or not. Moreover, the islanding boundaries are identified using an MILP optimization model based on linearized AC power flow. The proposed algorithm is tested on the IEEE 39-bus New England test system to reveal the performance and benefits.