Journal of Guidance Control and Dynamics · 1997 · 179 citations · 6 references
EngineeringAerospace EngineeringFlight DynamicsFlight MechanicsMachine LearningAdaptive ControlAircraft ControlControl TheoryManeuveringManeuvering FlightNeural Control SystemsAdaptive Flight ControlAdaptive NetworksAdaptive Neural ControlControl ArchitectureFlight Control ArchitectureFlight Dynamics ModelingNonlinear Aircraft DynamicsNonlinear Flight Control
Online adaptive neural networks are designed for nonlinear helicopter flight control architectures. The study focuses on the network architecture and examines how varying the adaptation gain influences control performance. Using a six‑degree‑of‑freedom nonlinear evaluation model of an attack helicopter, the network achieves near‑perfect tracking, cancels model inversion errors after brief learning, and its performance improves monotonically with higher adaptation gain while keeping control effort modest during the hover‑to‑forward transition.
The design of online adaptive neural networks for use in a nonlinear helicopter e ight control architecture is treated. Emphasis is given to network architecture and the effect that varying the adaptation gain has on performance. Conclusionsarebasedon asix-degree-of-freedom nonlinearevaluation model ofan attackhelicopter and ametricthatmeasuresthenetwork’ sability tocancel theeffectofmodeling errorsforacomplicated maneuver. Thenetwork isshownto providenearlyperfecttracking in thefaceofsignie cantmodelingerrorsand,additionally, to cancel the model inversion error after a short initial period of learning. Furthermore, it is shown that the performance varies gracefully and monotonically improves as the adaptation gain parameter is increased. The effect on control effort is modest and is mainly perceptible only during a short training episode that can be associated with transition from hover to forward e ight.
6