Adaptive Flutter Suppression for Aircraft Upset and Damage Conditions

Abstract

This paper presents an investigation of robust adaptive control for on-line aircraft utter suppression under upset and damage conditions, where two approaches are examined. The rst approach is based on generalized predictive control beginning by identifying directly the multi-step output prediction matrices under various test conditions from online timedomain input/output data. Singular value decomposition is then used to characterize and quantify the parameter uncertainties of the prediction matrices. A damage detection technique is developed to identify the damage status, type, and intensity. The predicted response error due to the parameter uncertainties and the error due to the identi ed damage are integrated into the predictive control design process for the performance of robustness and control under damage. The second approach is based on the integration of the real-time system identi cation with the associated uncertainty quanti cation and robust control design based on -synthesis, where the identi ed interval transfer function or the identi ed interval state space model is used for robust control design. The investigation is demonstrated by application to the Benchmark Active Controls Technology (BACT) wind-tunnel model.