Title :
A new interpretation of the LQG/LTR technique using optimal projection equations
Author :
Paschal, Randall N. ; Maybeck, Peter S.
Author_Institution :
Dept. of Electr. & Comput. Eng., Air Force Inst. of Technol., Wright-Patterson AFB, OH, USA
Abstract :
It is demonstrated that the LQR/LTR (linear quadratic Gaussian/loop transfer recovery) technique can be viewed as a way to achieve robustness even under the constraint of a reduced-order controller, even though one is not necessarily recovering a desired transfer function asymptotically. The optimal projection equation (OPE) approach gives an expanded view of LQG/LTR technique when the order of the controller is intentionally less than the order of the system design model. Also, the OPE approach allows other forms for Ω, which may give more flexibility as to how the system perturbations are modeled, to be chosen
Keywords :
optimal control; stability; LQG/LTR technique; linear quadratic Gaussian/loop transfer recovery; optimal control; optimal projection equations; reduced-order controller; robustness; stability; Control systems; Cost function; Filtering theory; Gaussian noise; Hilbert space; Noise measurement; Optimal control; Q measurement; Riccati equations; Robust control;
Conference_Titel :
Decision and Control, 1989., Proceedings of the 28th IEEE Conference on
Conference_Location :
Tampa, FL
DOI :
10.1109/CDC.1989.70163