MOSELM approach for Voltage Stability Indicator using phasor measurement units

Abstract

Voltage stability assessment is important in order to ensure a stable power system. Two algorithms were discussed in this paper which looks into estimating voltage stability based upon Thevenin Equivalent values in a system using Voltage and Current Phasors for different loading values. The first algorithm uses a Kalman filter based formulation. The second method uses an Online Learning approach known as the Modified Online Sequence Extreme Learning Machine (MOSELM). Results show that the Kalman Filter approach is capable of analyzing voltage stability but it requires some user specified information for tuning. On the other hand, the MOSELM approach show that it is capable of producing the same result as the Kalman Filter approach but require less amount of user specified information. © 2012 IEEE.