Nonlinear Output Feedback Control: Performance and Robustness

Michigan State University, East Lansing MI

Investigators

Abstract

ECCS-0725165 Khalil Despite some remarkable progress in nonlinear control theory, most of the available design methods focus on the asymptotic behavior (stability or regulation) and robustness. Very little analytical methods are available for shaping the transient response, which is essential for meeting control objectives. Intellectual Merit: Performance-recovery tools will be developed to recover the transient performance achievable in nonlinear control. High-gain observers with bounded control allow us to recover with output feedback the transient response achievable with state feedback. Conditional servo -compensators allow us to recover with servo-compensators the transient performance achievable without servo action, but with the steady-state accuracy of servo action. Extended high-gain observers, acting as disturbance estimators, allow us to recover in the presence of disturbance the transient performance achievable with known nonlinear models. The proposed work will develop the design tools, derive the theory that backs up the methodology, provide rules for tuning the design parameters, and demonstrate the effectiveness of the tools through experimental test-beds in electric drives and smart material-actuated systems. Broader Impacts: The research will promote teaching, training and learning by including Ph.D., M.S., and undergraduate students in the research activities, and by integrating research into control education through textbook writing. It will broaden participation of underrepresented groups.