Evaluación de Coherencia en el Sistema Nacional Interconectado Ecuatoriano empleando Mediciones Sincrofasoriales

Abstract

This paper presents the coherency evaluation of the dynamic coherency in the Ecuadorian power system, using measurements obtained from Synchrophasor Measurement Units (PMUs) located at the main substations of the backbone transmission system.

Based on the time series delivered by the PMUs, such as frequency, rate of change of frequency, voltage or current phasors, data mining techniques are applied to determine the coherent areas. Several batch clustering techniques has been applied to analyze post-contingency signals in an offline environment. For real time analysis, new algorithms based on stream data concepts have been implemented.

For the evaluation of coherent areas using post-contingency data, validation indexes are used to measure the quality of the clustering process, these indexes are calculated based on the density of the data within the same group or the distance between different groups. For the real time coherency analysis, recursive clustering algorithms are studied and implemented. These algorithms are capable of identifying the number of groups, as well as their composition, through the change of its parameters according to the last received information.