Análise da estabilidade e controle de conversores CC-CC conectados em cascata

Abstract

The need to use renewable energy sources has driven the use of distributed energy generation systems. This type of system is formed by several interconnected stages that can consist of filters, converters, inverters, among others. These elements are typically individually designed to ensure stability and performance. However, the interactions between the connected stages can lead the whole system to instability, which is physically characterized by oscillations that arise in the input and output buses. In the literature, most studies relate this problem to the Constant Power Load behavior of converters operating in closed loop. In addition, the number of studies that explain in detail the origin of these oscillations is scarce. Thus, this Thesis presents a methodology to analyze the stability of a system formed by cascaded DC-DC converters and to identify the source of possible oscillations. With the proposed methodology it was discovered that the stability problem is related to uncertainties present in the nominal model that represents the plant. Then, a maner is presented to identify the uncertainties through the Extra Element Theorem (TEE) and also through an analysis of parametric variations. From the analyzes carried out, the Model Reference Adaptive Control (MRAC) was chosen to solve the stability problem, since such an approach has the capacity to deal with uncertainties. It was shown that the MRAC controller was able to guarantee system stability with acceptable performance. However, in order to improve the plant's response, a Hybrid controller was proposed, composed of the combination of PID and MRAC techniques, with which it was possible to guarantee stability as well as improve performance. The whole theory presented is validated with simulation and practical results when applied in a system consisting of an LC input filter connected in cascade with a closed loop Buck converter.