Részletes program:

10:00 – 11:30  1st lecture block (Seminar I)

11:30 – 12:30  Lunch, Laboratory visit (Ebédszünet, laborlátogatás)

12:30 – 14:00  2nd lecture block (Seminar II)

Abstract:

Robust control theory explicitly deals with model uncertainty in its approach to the analysis and synthesis of closed-loop controllers. Robustness methods aim to achieve and/or analyze robust performance and/or stability in the presence of bounded modeling errors. Robust control theory dates back to Bode in the 1950s, with Horowitz expanding these ideas in the frequency domain during the early 1960s. The late 1960s and early 1970s saw to introduction of new state-space and frequency domain robust control techniques which capture the effect of model uncertainty on the closed-loop system. The introduction of robust control software tools in 1989 provided the academic and industrial community with access to control analysis and synthesis methodologies that capture the issues of model uncertainty, stability, performance, and the tradeoff between them.

The first seminar, “Robust Control: The Big Picture” provides an overview of robust control theory and its application to a flight control example. The robust control objectives are presented as are their incorporation into the analysis and synthesis of multivariable controllers. Three multivariable controllers are analyzed and compared. The results show the benefit of using the robust control framework to analyze and synthesize control algorithms.

The second seminar, “Robust Control: Tools of the Trade” provides an overview of the software tools available and their application to an automotive suspension control system. These tools allow inclusion of model uncertainty into the design process and provide methods to synthesize controllers which optimally tradeoff between robustness and performance objectives. The impact of model uncertainty on the control system performance and robustness is characterized and the worst-case combinations of uncertain elements identified. The control designs are validated by calculating worst-case gain and phase margins and worst-case sensitivity to disturbances.