Full text
| |
| Author(s): |
Luciano Antonio Frezzatto Santos
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | Campinas, SP. |
| Institution: | Universidade Estadual de Campinas (UNICAMP). Faculdade de Engenharia Elétrica e de Computação |
| Defense date: | 2017-07-31 |
| Examining board members: |
Pedro Luis Dias Peres;
Eugênio de Bona Castelan Neto;
Fernando de Oliveira Souza;
João Bosco Ribeiro do Val;
Matheus Souza
|
| Advisor: | Pedro Luis Dias Peres |
| Abstract | |
This thesis is concerned with the problems of robust static state and output feedback control synthesis and of robust filtering design for uncertain linear discrete time-invariant systems by means of dynamics enhancement. For that, the control laws are supposed to contain past information and the filter structure has a dimension greater than the dimension of the system being filtered. Extensions to cope with the smoothing problem are proposed as well. The procedure for the design of controllers and filters is performed by defining an auxiliary augmented system, from which convex synthesis conditions are derived, being described in terms of parameter-dependent Linear Matrix Inequalities. The H-2 and H-infinity norms are used as performance criteria for the developed methodology. The proposed conditions explore the particular structure of the auxiliary augmented system, thus being distinct from classical conditions from the literature for robust full-order filtering design, or for determining the state- or the output-feedback gains for the augmented system, in terms solely of the current instant. It is shown that the proposed methods encompass other techniques available in the literature, not incurring, therefore, in a greater degree of conservatism than the existing approaches. Numerical examples illustrate the advantages of the proposed conditions in determining lower upper-bounds for the H-2 and H-infinity norms, showing that the method tends to be less conservative as more past information is added to the control laws or as the order of the filters increases. Furthermore, the numerical examples evince that the same results are not attained by simply applying standard approaches from the literature to the auxiliary augmented system (AU) | |
| FAPESP's process: | 13/05957-1 - Stability and control of dynamical systems: uncertainty, constraints and polynomial models |
| Grantee: | Luciano Antonio Frezzatto Santos |
| Support Opportunities: | Scholarships in Brazil - Doctorate |