The paper investigates problems of designing controllers to linear time-invariant continuous descriptor systems subject to norm-bounded structured uncertainty so that the closed-loop systems are admissible or D-admissible with their transfer matrices having H-infinite norm bounded by a prescribed value. The constant state feedback and the dynamic output feedback designs are addressed. In both design methods, sufficient LMI conditions are derived to guarantee achievement of the desired specifications, such as robust H-infinite norm and pole-clustering constraints. Finally, two numerical examples are shown for the illustration.

This thesis presents strict LMI conditions for the bounded real lemma of discrete descriptor systems. Compared with existing nonstrict LMI conditions, the proposed new conditions are more tractable and reliable in numerical computations, in the sense that they can be tested easily by using the LMI Control Toolbox of Matlab. Based on the strict LMI conditions, the state feedback design for H-infinite control problem is also addressed. A sufficient LMI condition is derived so that the constructed feedback gain matrix from its solution will meet the design criteria of the closed-loop systems. Furthermore, we can probe into the problems of robust H-infinite control and pole-clustering in a disk for uncertain discrete descriptor systems subject to time-invariant norm-bounded uncertainty and convex polytopic uncertainty in the state matrix, respectively. Some sufficient LMI conditions are derived for analysis and design of these problems as well. Numerical examples are included to illustrate the results.

Orientador: Alberto Luiz Serpa / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-08-06T08:29:26Z (GMT). No. of bitstreams: 1 SarraciniJunior_Fernando_M.pdf: 6793847 bytes, checksum: effab2db853f18298e45c0a677c2ee50 (MD5) Previous issue date: 2006 / Resumo: A implementação de controladores de ordem reduzida (fixa) demanda um menor esforço de processamento e consequentemente recursos de hardware menos sofisticados em relação à implementação de controladores de ordem completa. Este trabalho mostra que a implementação prática de controladores H 00 de ordem fixa voltados para o controle de estruturas flexíveis é factível. A obtenção de tais controladores é um problema considerado difícil por ser nãoconvexo. Para contornar as dificuldades numéricas de obtenção dos controladores de ordem fixa, uma combinação do método Lagrangiano Aumentado com Desigualdades Matriciais Lineares (LMls) é utilizada. Uma estrutura de viga com engaste em uma de suas extremidades é modelada através do método de Elementos Finitos. Controladores Hoc de ordem fixa e de ordem completa são projetados com base em um modelo matemático truncado. Incertezas de modelagem e a presença de modos próximos na região de frequência de interesse dificultam a obtenção de controladores que garantam a estabilidade e um desempenho satisfatório. Para contornar estas dificuldades, usa-se a técnica de controle robusto Hoo e filtros de ponderação. Dessa forma, procura-se minimizar o efeito das incertezas e evitar que modos que não foram considerados durante a fase de projeto dos controladores não sejam excitados, garantido assim a não ocorrência do fenômeno denominado spillover. Controladores Hoo de ordem completa e ordem fixa são implementados na prática e os resultados experimentais são comparados com resultados simulados / Abstract: The implementation of reduced (fixed ) order controllers requires a smaller computational effort and. consequently, less advanced hardware resources in relation to the implementation of full order controllers. This work shows that the practical implementation of fixed order Hoo controllers directed toward the control of flexible structures is viable. Obtaining such controllers is considered a difficult task for being a non-convex problem. To overcome the numerical difficulties of attainment of fixed order controllers, a combination of the Lagrangian method increased with Linear Matrix Inequalities (LMIs) is used. A cantilever beam is modelled with the Finite Element Method. Fixed and full order controllers are designed based on a truncated mathematical model. Modelling uncertainties and the existence of near modes in the frequency range of interest make difficult the attainment of controllers that assure the stability and the performance of the system. To overcome this difficulty, the robust Hoo control and weighing filters are used. In this way, it is desired to minimize the effect of uncertainties and avoid the excitement of non-modelled modes, assuring that the spillover phenomenon does not occur. Full order and fixed order H x controllers are implemented in the practice and the experimental results are compared with the simulated results / Mestrado / Mecanica dos Sólidos e Projeto Mecanico / Mestre em Engenharia Mecânica