Identification of pilot nodes for secondary voltage control using K-means clustering algorithm

Abstract

The voltage control represents one of the most important qualitative factors, and it is addressed to the secondary power system control level. The performance of secondary voltage control depends strongly on the selection of pilot nodes. This paper proposes the use of K-means clustering algorithm order to the identification of pilot nodes for Secondary Voltage Control (SVC) in the power grids. In the first phase, the algorithm is based on the zoning of power grid into coherent and completely connected clusters (control areas). In the second phase, a node will be selected as pilot node for each control area. The pilot node must satisfy a minimum criterion based on the sum of electrical distances between it and all the others nodes. The obtained results were compared with other classical methods using a test system. The conclusion was that the algorithm can be used successfully in the optimal zoning of power grids and the identification of pilot nodes for SVC.