Multivariable strategy for modeling and control design of voltage-sourced and back-to-back converters

Abstract

The world’s energy system is currently undergoing a major transition in which power eletronic converters gain more importance. In this context, voltage-sourced converters just like their association on Back-To-Back converters, are rather common and have gained more relevance. This master thesis then presents the mathematical modeling of both converters in a multivariable representation strategy. Two test systems are designed: one system composed of a voltage-sourced conveter connecting an AC grid to a DC bus, in which a DC current source is used to represent the power profile of a DC microgrid, and a system composed of a BTB converter that interconnects two AC systems. For the first system, two multivariable control strategies are presented: a linear quadratic regulator (LQR) and a robust controller which is based on pole placement via linear matrix inequalities. Both controllers are designed so the converter might deal with a bidirectional power flow. For the second test system, one single controller is designed and it is based on pole placement via linear matrix inequalities. In this case, the controller is designed so the BTB converter might deal with bidirectional power flow, while being resilient against uncertainties and disturbances in the parameters that compose the AC system connected to one of the converter’s terminals. Both systems are implemented in a digital simulation environment and their dynamic responses are assessed and compared