No abstract

No abstract

This paper presents the model and the control of the rotor current loop a double-fed induction generator based on a robust generalized predictive controller. The controller is designed to minimize the control signal and ensure the reference tracking quickly and with a minimum of overshoot. Furthermore, it is proposed a new methodology for controller tuning based on a single parameter that allows a balance between noise attenuation of the control signal, disturbance rejection and robustness of the system during operation. Through computer simulation, the relationship between the tuning parameter and the system dynamics is analyzed. The simulations results validate the performance of the proposed controller when applied to a DFIG-based wind generation system. I.

Resumo -Este artigo apresenta uma nova estratégia de controle preditivo contínuo aplicado à malha de corrente elétrica do rotor de um sistema de geração eólico baseado no gerador de indução duplamente alimentado. O projeto do controlador é realizado a partir do modelo matemático da malha de corrente do rotor e sua sintonia é efetuada de forma simplificada através de dois parâmetros de ajuste, um para obter uma resposta rápida no seguimento de referência e o outro para reduzir as oscilações no sinal de controle. Desta forma, diminuindo oscilações no conjugado eletromagnético e nas potências instantâneas produzidas pelo sistema. Além disso, propõe-se a inclusão do modelo da saturação no controlador para evitar problemas de windup. Resultados experimentais são obtidos utilizando um protótipo de 3,7 kW que mostram a eficácia do controlador desenvolvido quando o sistema é submetido a variações de potência, velocidade do rotor e afundamento de tensão simétrico. propõem o uso do GPC (Controlador Preditivo Generalizado) para controlar o ângulo de passo das pás de uma turbina para reduzir as oscilações de potência mecânica e extrair a máxima potência de uma turbina eólica baseada no horizonte de predição da velocidade do vento como um parâmetro de desempenho. Devido à complexidade da topologia de controle e a quantidade de variáveis a serem controladas, Eletrôn. Potên., Campo Grande, v. 22, n. Palavras-Chave -

No abstract

No abstract

The use of windmills for the pumping of well water has appeared economically feasible in regions with average winds above 5 m/s, and when the water is used for human consumption and irrigation in arid rural zones, the importance of extracting the water from these wells efficiently and sustainably becomes more evident. However, concerning the state of Ceará/Brazil, only 18% of the drilled wells are freshwater wells, thus requiring additional energy for the utilization of desalination systems such as reverse osmosis. For that, the correct sizing of the pumping system associated with water desalination systems is essential, bearing in mind that the knowledge of the performance curves of the windmills, usually associated with piston pumps, is the "key factor" of the project. A brief analysis of the literature and websites of windmill suppliers shows that the performance curves of windmills associated with piston pumps are scarce, and the procedure to obtain them is not clear. This work's novelty is to use the windmill's power coefficient curve in association with the piston pumps to obtain the windpump performance curves, as pressure head versus volumetric flow rate in different wind conditions and sizes of windmills. The obtained curves were compared with other two curves found in the literature: one in a supplier's catalog and another from the "rule of thumb". The results showed a considerable divergence when calculating the root mean square deviation: 43.73% in relation to the supplier data average, and 125.76% in relation to the rule of thumb average results; however, the model proposed herein presented a limited operational range, which was not verified in the other two curves, which reproduced unlimited machines in terms of provided pressure head and flow rate. The unreal and asymptotic characteristics are more evident in the curves obtained by the rule of thumb, where the differences are also greater (125.76%) in relation to the proposed model. In terms of freshwater production, the model proposed herein reached, with an average wind speed of 6.5 m/s, and with a windmill of 6.1 m in diameter, a flow rate of up to 9.1 m 3 /day, after going through the reverse osmosis system. The same condition with the rule of thumb's model reached a flow rate of up to 11.3 m 3 /day. Finally, the model proposed herein for obtaining windmill's performance curves proved to be easy to implement and with realistic results, becoming an important design tool.

No abstract