Coordinated tuning of power system controllers using parallel genetic algorithms

Abstract

This work proposes a computational methodology for coordinated tuning of the power system controllers based on Genetic Algorithm which acts maximizing simultaneously two objective functions, representing each of them the damping of electromechanical oscillations and the improvement of the automatic voltage regulator responses, considering at the same time several critical operating conditions. The coordinated tuning procedure was posted as a multi-objective optimization problem, through the weighted sum technique. The controllers considered into the adjusting scheme correspond to the automatic voltage regulators, power system stabilizers and static var compensators to enhance the electromechanical and voltage response for a several critical operating conditions considered. The Genetic Algorithm was adapted for parallel computing in order to achieve both dynamic responses encompassing several critical operating conditions to reduce high computational efforts. Simulation results of the parallel implementation were more significant than the sequential version, and the proposed approach becomes an interesting alternative tool for operation planning and stability studies.