The problem of accurate tracking to a non-zero reference input is addressed in the context of multivariable flight control. There exists a requirement of zero steady state error in order to perform specific autopilot commands such as accurate 4-dimensional control: thus enabling the aircraft to be accurately positioned in time and space. The novel algorithm of optimal dyadic pole placement (ODP) is presented as the solution to this servomechanism problem. In addition, in order to propose a pragmatic solution, a two-stage control strategy is suggested which is shown to offer robust performance which may be implemented through a gain scheduling methodology. The two-stage control strategy achieves zero steady state error through the application of LQR to obtain an approximate track to the reference signal followed by ODP to effect the negation of the steady state error. The effectiveness of this approach is demonstrated via simulation. The vehicle considered is the Raven 2 remotely piloted vehicle.
展开▼
