This paper presents the "Adaptive Control Technology for Safe Flight (ACTS)" architecture, which consists of a non-adaptive controller that provides satisfactory performance under nominal flying conditions, and an adaptive controller that provides robustness under off nominal ones. The design and implementation procedures of both controllers are presented. The aim of these procedures, which encompass both theoretical and practical considerations, is to develop a controller suitable for flight.

more »

... he ACTS architecture is applied to the Generic Transport Model developed by NASA-Langley Research Center. The GTM is a dynamically scaled test model of a transport aircraft for which a flight-test article and a high-fidelity simulation are available. The nominal controller at the core of the ACTS architecture has a multivariable LQR-PI structure while the adaptive one has a direct, model reference structure. The main control surfaces as well as the throttles are used as control inputs. The inclusion of the latter alleviates the pilot's workload by eliminating the need for cancelling the pitch coupling generated by changes in thrust. Furthermore, the independent usage of the throttles by the adaptive controller enables their use for attitude control. Advantages and potential drawbacks of adaptation are demonstrated by performing high fidelity simulations of a flight-validated controller and of its adaptive augmentation. must be cancelled by the elevators. Auto throttle designs that only depend on the aircraft velocity rely on the pilot's ability to generate a suitable set of pitch commands to attain the desired cancellation. The ACTS controller pursues this cancellation automatically, thereby considerably reducing the pilot's workload. Simulation studies are used to illustrate some advantages and liabilities of adaptation. While these simulations only give a local notion of the performance and robustness characteristics of both the nominal and adaptive controllers, the companion paper [3] evaluates these characteristics from a more global perspective [2, 4] . Developments that enable tuning the control parameters to improve the resulting system's robustness are also presented. This paper is organized as follows. Section II presents the control structure of both the baseline and adaptive controllers. This is followed by Section III where a departure in the controller's implementation from the LTI framework supporting the adaptive control design procedure is made. Section IV demonstrates some advantages and potential drawbacks of adaptation via simulation. Finally, Section V presents conclusions.