Algoritmo genético para sintonia de controladores de sistemas elétricos de potência considerando o critério N-1

Abstract

Electrical Power Systems are often subject to electromechanical oscillations, which can be caused by a sudden increase of charge or even a short circuit that may lead to a transmission lines malfunctioning, avoiding a full quality distribution of electrical energy. In order to make such problems become less frequent or even extinct, power controllers have been used to enhance the system’s stability margin and, therefore, guarantying that such does not get into a collapse. Among countless power controllers, all used in industry for electromechanical oscillations damping, PSS (Power System Stabilizer) must be highlighted due to its capacity of increasing the stability margin of a SEP (Electrical Power System). The PSS has been used for decades due to the good performance associate its easiness in using their control techniques and that is why it is going to be applied in this assignment. However, only adding Power Systems Stabilizers does not guarantee the required stability of an electrical power system that is why is necessary to have a proper alignment of these two in order to turn the system robust enough against the variations in its operation points. This proposal presents an alternative automatic tuning for the PSS (Power System Stabilizer) type of controllers from electrical power systems, using a Matlab® based genetic algorithm and considering the N-1 criterion established by the ONS (Operador Nacional do Sistema). In order to increase the challenge of this assignment and to test the endurance of the proposed algorithm, the used testing-system Will be “New England-New York”, containing 68 bus and 16 generators. Intending to promote a situation in which the controllability would be more difficult, the lines are removed, one by one, from the electrical system, as previewed by the N-1 Criterion, being each case originated with such methodology a distinct operation point to be considered in the project. Therefore, the proposed genetic algorithm will have as main goal to find the adequate standards for all the controllers, used in order to synch simoutaneously all the considered operation points, that is, the parameters found by the algorithm must be robust enough to deaden the electromechanical oscillations of the system in every operation point, until the stopping criterion established by the projector is reached (usually a minimum deaden value for the system). The main contribution of this assignment is to use the N-1 Criterion, established by the ONS, to verify the automatically synched controllers performance by the algorithm proposed.