Controle robusto multivariável aplicado ao sistema de geração fotovoltaico conectado à rede

Abstract

This work presents a study of the photovoltaic converter from a multivariable perspective. The mathematical model of the system is developed using the state-space method, followed by its linearization at points of interest, in order to obtain the transfer function matrix. Frequency response analysis of the converter is conducted based on singular value decomposition and relative gain array techniques. Singular value decomposition is employed to identify the directions in which a non-square system cannot be controlled, while relative gain array frequency response analysis allows determining which variables are more suitable for control over others. A centralized robust control strategy is proposed, with feedback gains calculated by solving linear matrix inequalities. This approach considers parametric variations of the converter resulting from changes in power supplied by the photovoltaic array. To mitigate the influence of disturbances on the converter’s operation, an optimization based on the H∞ norm is used, implemented through linear matrix inequalities. The performance of the control structure is evaluated through simulations conducted in the PSIM software. The centralized control loop was able to meet the design requirements for settling time and maximum overshoot. Additionally, the multivariable controller achieved decoupling of the state variables through the application of feedback gains.