Alocação computacional inteligente de autoestruturas para controle multivariavel

AUTOR(ES)
DATA DE PUBLICAÇÃO

2000

RESUMO

In this thesis a proposal for eigenstructure placement for multivariable linear dynamic systems by state feedback based on the linear quadratic regulator (LQR), multiobjective optimization, evolutionary computation and parallel programming is presented. The eigenstructure placement problem is formulated based on LQR designs and on an inequality method to conceive a multiobjective optimization problem for the proposed formulation. The solution to this problem is obtained via the development of a parallel genetic algorithm dedicated to the search of LQR design s weighting matrices. Search strategies are formulated and elements of computational intelligence are used to model them and their implementations are made on a logical decision unit that interacts with the GA-optimizer

ASSUNTO(S)

autovalores teoria do controle algoritmos geneticos analise modal processamento paralelo (computadores) sistemas de controle por realimentação controle automatico - sensibilidade

ACESSO AO ARTIGO

Documentos Relacionados