Comparison and Combination of Shuffled Frog-Leaping Algorithm and K-Means for Clustering of VCAs in Power System

Voltage Control Areas, is one of the main concepts in both voltage stability assessment and control. Employing an artificial intelligent system, for detecting VCAs and identifying the prone buses is an important part of a reasonable response to the need for real time monitoring and control of power system in emergency cases of voltage instability appearance. For this purpose the database of training data, contains the simulation results due to wide range of various emergency cases. A clustering process which provide finite clusters belong to VCAs, in addition of a classification function based on the provided clusters, are the main stages to prepare A IS for automatic VCA identification. In this paper a novel data clustering method based on Shuffled Frog Leaping Algorithm is presented to cluster the VCAs in excessive emergency cases in electric power system. In present study the application of SFLA in data clustering is also compared with the most popular analytic algorithm of clustering, K-means, and also the hybrid SFLA and K-means method are used for better clustering performance. The advantage of using the data points from the same under clustering database as initial positions for some frogs is illustrated too. Numerical results are presented for IEEE 14-bus test system and two standard test data sets. The comparative results show the effectiveness of proposed algorithms. Copyright (C) 2010 Praise Worthy Prize S.r.l. - All rights reserved.