Multivariable Control: Reduced-Order Observers

This lecture covers the design of reduced-order observers in multivariable control systems. The goal is to create observers with an order lower than the system's states, reconstructing the remaining states. Topics include system dynamics, change of coordinates, observer design, and examples like satellite attitude control. The lecture also discusses the design of full-order and reduced-order observers, estimator dynamics, and the assignment of dominant eigenvalues for closed-loop systems. Various design methods and heuristic approaches are presented, along with the importance of observers in systems with unmeasured states and the key role of eigenvalue assignment in output-feedback controller design.