Dynamic gain-scheduled control of the ICE 101-TV

Abstract

This paper shows the theoretical development and application of dynamic gain scheduled control -a novel method for the control of nonlinear systems -to an aircraft model. The idea behind this method is to schedule the control law gains with a fast varying state variable rather than with a slow varying state or an input parameter. This is advantageous as it is then possible to schedule the gains with a state variable that is dominant in the mode that we are most interested in controlling. The use of this type of gain scheduling is shown to improve the transient response of the aircraft model when stepping between trim conditions and to overcome some of the problems associated with conventional gain scheduled controllers (such as control surface position limit saturation). 'Hidden coupling terms' that introduce unwanted dynamics when scheduling gains with a fast state (rather than the input design parameter) are eliminated directly by applying a transformation to the classical parameter-scheduled gain distributions which are calculated using eigenstructure assignment. A second order longitudinal model and a 5th order longitudinal/lateral model of the ICE 101-TV tailless delta-wing aircraft configuration are used to demonstrate the design process. AbbreviationsAMT all-moving tip AMTact all-moving tip actuator AMTL all-moving tip (left) AMTLact all-moving tip (left) actuator THE AERONAUTICAL JOURNAL DECEMBER 2005 593

Two-state dynamic gain scheduling control applied to an F16 aircraft model

“…DGS was applied successfully on the ICE 101-TV [5], a wind tunnel model approximately representing the BAe Hawk aircraft [6] and more recently to an F16 model [7,8]. In all these cases, the scheduling for each gain was with respect to one state variable (or one plus an input); DGS in this form was derived for the case where each feedback gain is scheduled in terms of the state variable signal in its feedback path (it was also extended to account for the scheduling state being different from the feedback signal).…”

Dynamic gain scheduled control of a satellite with a robot manipulator

“…However, it is well-known that simple continuous gain-scheduling of linear design may not retain stability. To counter this, dynamic gain scheduling (DGS) [10][11][12][13][14][15] is adopted.…”

“…Jones developed it for higher order 'real' aircraft models [11] [12]. Hidden coupling terms are avoided which in other methods such as 'static' gain scheduling (i.e.…”

Dynamic gain scheduled control of a Hawk scale model

“…Classical gain scheduled (CGS) controllers are generally restricted by a 'slow variation' condition which means that the designed stability is only maintained if the system moves relatively slowly between operating points. The method developed in previous work (2,3) and applied experimentally in this paper addresses this issue. A controller is designed which allows the scheduling variable to be one of the system states.…”

“…Previous work carried out at Bristol into dynamic gain scheduled controllers presented in (2,3,5,6) has only involved theoretical results and numerical models. In this paper, the method has been applied to a Hawk dynamic test rig and the experimental results presented clearly demonstrate the advantages of applying DGS to a real system.…”