Trajectory tracking with fault tolerant flight control system: a model predictive control approach

This thesis presents new methods of Fault-Tolerant Control (FTC) using Model Predictive Control (MPC) techniques for trajectory tracking. Previous work has applied a finite prediction horizon MPC to provide reconfigurability and fault accommodation in the case of an effector fault. This thesis proposes the use of infinite horizon MPC formulations and proposes two versions of Constrained Infinite Horizon Model Predictive Control (CIHMPC) for fault-tolerant control. First, the problem of controlling an aircraft solely with the engines is considered and a CIHMPC is proposed to steer the aircraft through a pre-determined path, defined by geodetic waypoints, showing satisfactory maneuverability and full exploitation of the propulsion system up to its limitations. Second, a novel implementation of CIHMPC is proposed where the response of a reference closed-loop model is followed even in the presence of effector faults (in general). The proposed architecture is capable of redistributing, in a stable manner, the control efforts among healthy effectors, respecting its pre-defined limitations. The problem of evaluating the reduced maneuverability of an aircraft with effector faults is also considered and polyhedral sets describing the degraded performance of the vehicle are obtained, in a description suitable for usage in higher level control loops. Moreover, the inner loop equivalent dynamic characteristics are determined through system identification techniques. A constrained guidance system is then proposed that considers the limitations of the inner loop as input constraints, in order to smooth the transition between two consecutive navigation legs defined by waypoints. The trajectory tracking system composed of the constrained guidance and model following CIHMPC, is demonstrated in flight test on the VFW-614 AT- TAS (Advanced Technologies Testing Aircraft System) with simulated aileron fault, demonstrating adequate performance at multiple levels of control and with proper intrinsic robustness.