@PhDThesis{Santos:2015:InFuTé,
author = "Santos, Severino Pinto dos",
title = "Investiga{\c{c}}{\~a}o da fus{\~a}o da t{\'e}cnica de controle
fuzzy e/ou sdre/h\∞ visando projetar um controlador robusto
para um sat{\'e}lite r{\'{\i}}gido-flex{\'{\i}}vel",
school = "Instituto Nacional de Pesquisas Espaciais (INPE)",
year = "2015",
address = "S{\~a}o Jos{\'e} dos Campos",
month = "2015-05-28",
keywords = "H\∞,, control, controle Fuzzy, controle de atitude, controle
SDRE, controle de algoritmos gen{\'e}ticos, H\∞,, control,
Fuzzy control, attitude control, SDRE control, genetic algorithms
control.",
abstract = "Consultando-se a literatura sobre controle de sistemas,
constata-se o n{\'u}mero sempre crescente de projetos de controle
usando as t{\'e}cnicas de controle robusto, dentre as quais se
destacam as t{\'e}cnicas LQR (controle linear quadr{\'a}tico),
H\$_{INFINITO}\$. e SDRE (equa{\c{c}}{\~a}o de Ricatti
dependente do estado). O emprego destas t{\'e}cnicas na {\'a}rea
aeroespacial {\'e} cada vez maior, principalmente para controle
de sat{\'e}lites e plataformas espaciais. Os projetos de controle
robusto usando as t{\'e}cnicas LQR, SDRE e H\$_{INFINITO}\$,
atualmente, apresentam o inconveniente que consiste na trabalhosa
determina{\c{c}}{\~a}o por tentativa das matrizes de
pondera{\c{c}}{\~a}o destas t{\'e}cnicas, sendo estas matrizes:
Q e R para as t{\'e}cnicas LQR e SDRE e as matrizes W\$_{S}\$ e
W\$_{C}\$ para a t{\'e}cnica H\$_{INFINITO}\$. Este trabalho
prop{\~o}e minimizar o inconveniente citado acima, investigando e
projetando controladores multivari{\'a}veis para o controle e
estabiliza{\c{c}}{\~a}o de atitude sat{\'e}lites, usando a
fus{\~a}o das t{\'e}cnicas de controle \emph{Fuzzy} e/ou
algoritmos gen{\'e}ticos com as t{\'e}cnicas de controle LQR,
SDRE e H\$_{INFINITO}\$. Inicialmente s{\~a}o feitas
revis{\~o}es de um modelo de sat{\'e}lite r{\'{\i}}gido,
representado pelo simulador de sat{\'e}lite constru{\'{\i}}do
pelo INPE (Instituto Nacional de Pesquisas Espaciais) e de um
modelo de sat{\'e}lite r{\'{\i}}gido-flex{\'{\i}}vel,
representado pelo dispositivo SRV2, fornecido pela
Quanser\$^{®}\$, que consiste de um cubo no qual est{\'a}
engastada uma haste flex{\'{\i}}vel. Em seguida, usando a
t{\'e}cnica de controle \emph{Fuzzy}, obt{\^e}m-se de forma
autom{\'a}tica a matriz Q constante, para os controladores LQR e
SDRE, a matriz R {\'e} mantida como uma matriz diagonal
unit{\'a}ria. Finalmente, usando as t{\'e}cnicas de algoritmos
gen{\'e}ticos, s{\~a}o determinadas as matrizes de
pondera{\c{c}}{\~a}o de sensibilidade e de sensibilidade
complementar, respectivamente W\$_{S}\$ e W\$_{C}\$, de um
controlador H\$_{INFINITO}\$ sobre o sistema
r{\'{\i}}gido-flex{\'{\i}}vel. ABSTRACT: Referring to the
literature concerning on control systems, it can be seen un
increasing number of control designs using robust control
techniques, among which stands out techniques LQR (linear
quadratic control), H\$_{INFINITY}\$ and SDRE (state-dependent
Ricatti equation). The use of these techniques in the aerospace
area is increasing, mainly to control satellites and spatial
base.The robust control designs using the LQR techniques, SDRE and
H\$_{INFINITY}\$ currently have the disadvantage that is the
laborious determination by trial of the weighting matrices of
these techniques, and the Q and R matrices for the LQR and SDRE
techniques and W\$_{S}\$ matrices and W\$_{C}\$ for
H\$_{INFINITY}\$ technique. This paper proposes minimize the
inconvenience cited above, investigating and designing
multivariable controllers for control and stabilization of
satellites attitude, using the fusion of \emph{Fuzzy} control
techniques and / or genetic algorithms with the LQR control
techniques, and SDRE H\$_{INFINITY}\$. Initially revisions are
made of a rigid satellite model, represented by the satellite
simulator built by INPE (National Institute for Space Research)
and a rigid-flexible satellite model, represented by SRV2 device,
provided by Quanser\$^{®}\$, consisting of a cube in which is
embedded a flexible stem. Then, using \emph{Fuzzy} control
technique, one obtains automatically a constant matrix Q for the
LQR and SDRE controllers. The R matrix is maintained as a unitary
diagonal matrix (R=1) Finally, using the techniques of genetic
algorithms, determine the sensitivity and complementary
sensitivity weighting matrices, respectively W\$_{S}\$ and
W\$_{C}\$, for H\$_{INFINITY}\$ controller for a
rigid-flexible system.",
committee = "Ricci, Mario Cesar (presidente) and Souza, Luiz Carlos Gadelha de
(orientador) and Pilchowski, Hans-Ulrich and Morimoto, Nilton
Itiro and Pillat, Valdir Gil",
copyholder = "SID/SCD",
englishtitle = "Fusion research Fuzzy control and / or sdre / h\∞ aiming to
design a robust controller for a rigid and rigid-flexible
satellite",
language = "pt",
pages = "221",
ibi = "8JMKD3MGP3W34P/3JDEP6H",
url = "http://urlib.net/ibi/8JMKD3MGP3W34P/3JDEP6H",
targetfile = "publicacao.pdf",
urlaccessdate = "27 abr. 2024"
}