Dynamic gain scheduled control of a Hawk scale model

When designing flight control laws using linearisations of an aircraft model about different flight conditions, some form of scheduling of the resultant gains will often be required to implement the controller over wide operating regions. In practice, the controller gains are often scheduled against relatively slowly-varying system states such as altitude or velocity. However, it may also be desirable to schedule gains against rapidly-varying states such as angle-of-attach, thereby generating a cyclic dependence through hidden coupling terms. Previous published work at Bristol has developed a numerical method of accounting for this dependenc when scheduling state feedback gains against coupled states. The resulting 'dynamic gain schedule' is shown to significantly improve the transient response of the aircraft model during rapid manoeuvring and to reduce the chances of control surface actuator position limit saturation. In this paper, the novel design process, using eigenstructure assignment, is applied to a mathematical second-order longitudinal aircraft model which represents an approximate BAe Hawk wind-tunnel model. The dynamic gain scheduled controller is shown to work extremely well in practice when applied to the closed-loop experimental rig. Despite the highly nonlinear characteristics of the model aerodynamics and tailplane actuation system, as well as unmodelled high turbulence levels, dynamic gain scheduling demonstrates stable closed loop control even in regions where the nonlinearities are such that conventional gain scheduling fails to produce a stable response.