Application of Model Reference Adaptive Control to General Aviation

Abstract

In an effort to increase general aviation safety, model reference adaptive control (MRAC) is applied to a dynamic simulation of the Hawker Beechcraft CJ-144 fly-by-wire aircraft. MRAC controls the aircraft in the nominal case and adapts to maintain control during unexpected failures. An artificial neural network is trained online to correct for uncertainty and to adapt to changes or failures in the aircraft. The aircraft response is shaped by a model following linear controller. A time delay metric is used to tune the controller gains. Time response results are presented. This controller is designed to be flight tested on the CJ-144 aircraft.