Repositori URV

Identificador: TDX:1108

Handle: https://hdl.handle.net/20.500.11797/TDX1108

Autors:
Torres Pinzón, Carlos Andrés

Resum:
This thesis describes the theoretical background necessary to design and implement of analog robust control strategies based on Linear Matrix Inequalities (LMI). The control strategies proposed has been used in basic and cascaded of dc-dc converters topologies. The robust control methods can simplify and systematize the choice of feedback gains, taking into account the converters nonlinearities and uncertainties. These control methods ensure some prescribed restrictions on the pole placement and the control effort are met, and optimize the load disturbance rejection. The first method consists of two different robust controllers for dc-dc converters with two stages in cascade. The first robust controller is monovariable, this means, the duty-cycle is the same for the two switches, while the second robust controller is multivariable robust controller, this means, with independent duty cycles in each switch. Finally, the second method is a synthesis of fuzzy control based on LM I that has been adapted to bilinears basic dc-dc converters. Concretely, this technique is a linear-piecewise continuous control which optimizes the load disturbance rejection under pole placement constraints and duty-cycle limitation.