A new concept for retrofitting a reconfiguration module to an existing control law is reported in this thesis. The concept is motivated by the need for low cost, add-on modules that improve air safety in the existing fleet of civil air transport vehicles. A direct adaptive approach that accommodates control surface nonlinearities is adopted, which uses a slowly adapting model of the closed-loop aircraft as the reference model. The motivation, benefits, and components of the architecture are presented. In addition, the issues of control surface magnitude and rate saturation are addressed. A proof of stability is outlined for input-error adaptation when position and rate saturation are present. The reconfiguration architecture is demonstrated using an F/A-18 and a generic transport nonlinear simulator. General issues associated with commercial transport reconfiguration are highlighted. In both the longitudinal and directional axes, the control surfaces are not well balanced from a reconfiguration viewpoint. As a result, a novel reconfiguration control allocation scheme was devised that blends in all the control effectors in a given axis to perform the reconfiguration task. The simulation results revealed that the reconfiguration architecture does provide reconfiguration functionality for a wide variety of control surface failures. The reconfiguration potential is illustrated through comparisons of post-failure performance with and without reconfiguration via non-linear simulations. Additionally, comparisons between post-failure performance and nominal performance are made through non-linear simulations, closed-loop frequency responses, and aircraft handling qualities. For all of the failure scenarios illustrated, the simulation results showed that the aircraft without reconfiguration departs; with reconfiguration, nominal performance is achieve provided that adequate control authority exists post-failure.
展开▼
