ABSTRACT
Oregon State University, Department o f Electrical and Computer Engineering, Corvallis, OR 97331, USA
Nonlinear control of high-performance aircraft such as HARV (modified F-18) has been demonstrated with accurate computer simulations to be effective by our research [1] as well as others such as [2] and [3]. The difference between our controllers and the latter two is that the latter generate the nonlinear feedback gains to account for certain (assumed known) nonlinear plant dynamics, while our controllers do not involve the idea of linear approximation. Reference [2] utilizes numerous trim-state linearization studies to determine the required gains in conjunction with flight-tested controller which is able to successfully control HARV from a (to) = 5° trim to about 60° in about 5 seconds. Reference [3] is based on a linear Hoq design in conjunction with trim-state linearized dynamics and an
appropriate nonlinear gain scheduled according to dynamic pressure variation. The latter study only considers a maximum change in angle of attack (a) from a (ίο) = 10° to a = 20° in about 3 seconds with a rise time of 1 second. While neither of the latter two are nearly minimum-time maneuvers as demonstrated here, they probably represent the best controllers based primarily on linear design methodology in conjunction with somewhat ad-hoc nonlinear correction. Here we discuss synthesis of truly nonlinear controllers to obtain near-optimal performance of the aircraft. The model of the aircraft used in the simulations includes nonlinear aerodynamic coefficients and actuator constraints.
