Dynamic control allocation for reusable launch vehicles

Gain-scheduled linear-affine control allocation functions δ_c=F(x)τ_c+δ₀(x) are inherently inaccurate due to inherent limits of performance of the design process. As a result, online CA correction is required, either in terms of adaptive methods, e.g., system identification, or time-invariant methods, e.g., H_∞ control. We investigate the use of LTI compensator methods for dynamic control allocation (DCA). Design issues include closed loop performance and time-scale separation between the closed-loop vehicle dynamics, the autopilot, and the DCA compensator.