An extension of the linear quadratic Gaussian-loop transfer recovery procedure

Laleh Hosseini-Ravanbod (1) , Dominikus Noll (1) , Pierre Apkarian (2)

1 IMT - Institut de Mathématiques de Toulouse UMR5219
2 ONERA - The French Aerospace Lab [Toulouse]

The linear quadratic Gaussian-loop transfer recovery procedure is a classical method to desensibilise a system in closed loop with respect to disturbances and system uncertainty. Here an extension is discussed, which avoids the usual loss of performance in LTR, and which is also applicable for non-minimum phase systems. It is also shown how the idea can be extended to other control structures. In particular, it is shown how proportional integral derivative controllers can be desensibilised with this new approach. The method is tested on several examples, including in particular the lateral flight control of an F-16 aircraft.

Mots clés: LQG/LTR; observer-based control; PID control; mixed H 2 /H ∞ synthesis; trade-off; structured control law

Domaines: Mathématiques [math]; Optimisation et contrôle [math.OC]

Laleh Hosseini-Ravanbod, Dominikus Noll, Pierre Apkarian. An extension of the linear quadratic Gaussian-loop transfer recovery procedure. IET Control Theory and Applications, 2012, 6 (14), pp.2269 - 2278. ⟨10.1049/iet-cta.2011.0308⟩. ⟨hal-01868671⟩