An Intelligent Controlled Islanding Scheme for power systems

In this paper a new Intelligent Controlled Islanding Scheme (ICIS) able to change its behavior according to the state of the system is presented. As a first step, generators are classified into coherent groups by means of a recursive clustering algorithm applied to the measurements delivered by Phasor Measurement Units (PMU). In a second, tie-lines to be opened in order to create the islands are determined by means of concepts based on graph theory, shortest electrical distance and active power imbalance. A reduced frequency response model is employed to determine the power imbalance limits on each island. Finally, a DC power flow based procedure is used to assess and control possible overloads after islanding. The 39 bus - New England System is used to test proposed islanding scheme showing its capability to avoid power systems collapse.