Robust Adaptive Control of a Structurally Damaged Aircraft

Abstract

Application of adaptive control for aerospace applications has been a topic of great research interest over the past several decades. The driving force behind it is the enhanced safety that it can provide during adverse flight conditions and aircraft structural damage. Neural network based direct model reference adaptive controllers (MRAC) are particularly proven to be effective in handling such scenarios. But one of the limitations with this methodology is their inability to handle unmodelled dynamics, which severely restricts the adaptation rate of these controllers. This paper uses the recently developed modified state observer (MSO) adaptive control methodology which overcomes the above problem by using an observer controller structure that separates the design of estimation error dynamics from the nominal system dynamics. This allows for designing faster estimation error dynamics which permits the use of large adaptation rate. In this paper the MSO methodology is applied for the adaptive control of a structurally damaged aircraft, which has pitch ,yaw and roll axes dynamically coupled .So the adaptive control problem is to reduce the interactions between these axes and stabilize the aircraft. Simulation results demonstrate the potential of the proposed methodology in solving the above problem.