Sintonia automática e coordenada de controladores de sistema de potência utilizando algoritmo genético multiobjetivo

Abstract

The Power System is constantly subjected to disturbances, which is often a result of the unbalance between generation and consumption. These disturbances generate electromechanical oscillations and they can cause the instability of the system. To damp these oscillations, the control system called Power System Stabilizer (PSS) is added to Automatic Voltage Regulator (AVR) of the generators. However, its necessary tuning both AVR-PSS controllers in order to damp the oscillations and not influence the system dynamics. This automatic and coordinated tuning is commonly done using local or global search algorithm. In this work will be used a multiobjective global search algorithm, known as the NSGA-II, Non-Dominated Sorting Genetic Algorithm, because of its evolutionary characteristics that provides a good approximation of the optimal region for the solutions of the problem. A multiobjective optimization must be used because the problem under study in this work has conflicting objectives, maximization of damp of the mechanical oscillations of the system, through PSS, and maximization of AVR gain to improve the voltage regulation.