This thesis presents the experimental and simulated results obtained in the project and implementation of Generalized Predictive Control, an advanced digital predictive control technique, and fuzzy logic, applied to control and improvement of the stability of electric power systems. One of the main contributions of this thesis is of pratical nature, since a part of the proposed predictive control strategies were experimentaly avaliated by real avaliation tests. These tests are realized with a small scale real power system. There are other contributions like these: a) the project and validation of a computational power system simulator used to evaluate the behavior of the proposed control structures when operational contingencies, commonly observed in normal operation of the power system, are simulated; b) assembly a prototype of a small scale real electric power system used to evaluate a part of the proposed predictive control techniques. The assembled power system prototype is formed by two electric machines: a DC motor and a salient-pole synchronous generator. These two machines are mechanicaly coupled and connected to a commercial power system through a small transmission line simulator. The prototype was assembled at the Centro de Tecnologia da ELETROBRAS-ELETRONORTE (LACEN), located in Belém, Pará, Brazil. Using the prototype's features, it is possible to evaluate and validate advanced monitoring and control strategies. After these avaliations, these strategies can be used as predictive maintenance tools for the rotating electrical machines in operation at ELETROBRAS-ELETRONORTE.
|
