This paper provides a new design procedure for robust current controllers applied to LCL-filtered grid-tied inverters suitable for the integration of renewable energy sources. The design takes into account the digital implementation delay, multiple resonant controllers and operation under uncertain grid impedance. The procedure is based on the optimization of an objective function that allows to get a good trade-off between the settling of transient responses and rejection of disturbances. A particle swarm algorithm is used to find the optimal control gains and, differently from other works, here the robust stability of the closed-loop system under uncertain parameters is theoretically certified by means of linear matrix inequalities. Experimental results are shown, confirming that the closed-loop system with gains obtained by the proposed procedure presents a good tradeoff between robustness and performance, with suitable transients and grid currents with low harmonic content, complying with requirements from IEEE 1547 Standard, becoming a useful robust control design alternative for power converters in the distributed generation scenario.
Este trabalho apresenta o projeto de um controlador robusto de corrente por realimentação de estados baseado em desigualdades matriciais lineares, conduzindo a um conjunto de ganhos fixos capaz de garantir a alocação dos polos de malha fechada em um círculo de raio mínimo, centrado na origem e contido no círculo de raio unitário (controlador robusto quasi-deadbeat). Para aplicações em conversores conectados à rede, o procedimento de projeto proposto produz resultados ilustrando que relaxações a partir do controlador robusto quasi-deadbeat garantem estabilidade e bom desempenho para uma faixa de incertezas e para variações nos parâmetros da rede e do filtro.
This paper proposes a new design procedure for state feedback controllers suitable for practical application in grid-connected inverters, leading to results complying with IEEE 1547 Standard, even with a reduced number of sensors. The procedure relies on the optimization of objectives considering stability, tracking error and disturbance attenuation. A genetic algorithm is used to search, based on simple models, the control gains, leading to a set of viable solutions (Pareto front). The chosen control gain has robust stability against parameter uncertainties and variations certified by means of analysis linear matrix inequalities. Constraints in the search space allow to find full or partial state feedback Artigo submetido em 26/09/2018. Primeira revisão em 24/11/2018. Aceito para publicação em 17/01/2019 por recomendação do Editor Marcello Mezaroba. http://dx.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.