Neuro-adaptive approach for fading actuation accommodation in space flight vehicles

One of the typical yet devastating faults in flight systems Is the loss of the driving (propulsion) power. In this work, we explore an Innovative approach to accommodate such a fault of space flight vehicles operating under varying fight conditions. We present a neuro-adaptive control strategy, composed of two Neural Network (NN) units, the former to adapt to unanticipated uncertainties and the latter to enhance performance by compensating (funneling) the estimation error caused by the first NN to incorporate the Important and most common problem faced by the airborne space vehicles, unforeseen and uncertain actuating failures. New algorithms are derived to cope with the sub-system failures due to the jet engine partially losing its propulsion power ensuring the system stability. Simulation study using a generic model similar to X-37 flight vehicle model demonstrates a dramatically improved performance in the face of fading power faults and system uncertainties. Copyright © 2006 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.