Lateral imbalance detection on a UAV based on multiple models

This paper addresses a multiple-model based lateral imbalance detection methodology for an uninhabited air vehicle (UAV). Two critical imbalance failures are considered that are the failure-induced left aileron stuck and the centre-of-gravity shift along the y-axis. A bank of LTI Kalman filters are designed to detect the above lateral failures and a flight control law based on the model predictive control (MPC) theory is designed for the aircraft lateral directional dynamics. It is shown that the proposed multiple-model detection scheme is able to achieve an effective reconfiguration capability to provide the efficient handling qualities at the failure-free flight operating conditions whilst it maintains desirable performance at post-failure conditions. The results of the proposed multiple-model based fault reconfigurable scheme for the UAV flight dynamics are illustrated and validated through simulations.