Adaptive control of aircraft with unknown faults

Abstract:

Changes in aircraft dynamic model caused by body damage or other unknown faults may cause pre-designed control laws unsuitable for maintaining steady flight and adaptive control methods are needed to ensure flight safety. For this study, based on the nonlinear NASA Generic Transport Model (NASA_GTM), retrospective cost adaptive control (RCAC) is applied to maintain steady flight with unknown faults, including icing and control surfaces faults. With knowledge of Markov parameters which are calculated from the linearized version of NASA_GTM, output feedback RCAC controller is designed. Retrospective cost function is further optimized to update adaptive gains of RCAC controller and the thrust and control surfaces are adaptively modified to compensate for the impacts of unknown faults. The results show that RCAC controller is able to adapt its gains in order to maintain the steady flight and follow control commands.