Robust adaptive control of a highly nonlinear ill-conditioned plant

An adaptive procedure is proposed for the robust control of a highly nonlinear ill-conditioned plant - the Flying Test Bed (FTB): a rocket that will be used to carry an Air-Turbo Ram jet (ATR) engine during engine tests. Three models - a nonlinear model, a linear time-varying model with nonlinear gravity compensation and a linear model in the form of a transfer function - were developed. The third model is used for designing robust controllers. For airspeeds below 100m/s a fixed robust controller can be designed for large variations in all parameters; however, for higher rocket velocities the range of allowable parameter variations must be reduced in order to find a fixed controller. The second model is used for the identification of the rocket's uncertain parameters: C
_Lα
, C
_Lδ
and lcp. As they are nearly unidentifiable at lcp approximately equal to 0, i.e., close to the rocket's instability, a method running a set of parallel models - observers is proposed, and the nearest neighbor in the dynamic response is used as the resulting model. The first model is used to test the proposed procedure by the simulation of the rocket in a dive-recovery experiment. The simulation results proved the robustness of the proposed control scheme.