Fault-Tolerant Model Predictive Control Trajectory Tracking for a Quadcopter with 4 Faulty Actuators

Akram Eltrabyly (1) , Dalil Ichalal (1) , Saïd Mammar (1)

1 IBISC - Informatique, BioInformatique, Systèmes Complexes

In this paper, an active fault tolerant control (AFTC) framework is presented, with the ability to automatically estimate and compensate for the Loss Of Effectiveness (LOE) of all actuators of a quadcopter Unmanned Air Vehicle (UAV), while achieving trajectory tracking. The aforementioned AFTC framework consists of a reconfigurable controller based on Nonlinear Model Predictive Control (NMPC), and a fault detection and diagnosis (FDD) module based on a simple and computationally cheap nonlinear algebraic observer algorithm. The presented FDD module estimates the multiplicative faults in all quadcopter actuators simultaneously in an accurate and timely manner. A nonlinear simulation validates the proposed framework and proves a precise trajectory tracking performance in the presence of 4 simultaneous actuator faults. Furthermore, simulation results show that this framework is numerically tractable and real-time applicable, as the ratio of run-time to simulation time is much less than one.

Mots clés: Algebraic observer; Diagnosis; Fault detection; Fault estimation; Fault-tolerant control; Model predictive control; Multiplicative faults; Reconfigurable control

Domaines: Automatique / Robotique

Akram Eltrabyly, Dalil Ichalal, Saïd Mammar. Fault-Tolerant Model Predictive Control Trajectory Tracking for a Quadcopter with 4 Faulty Actuators. 4th IFAC Conference on Embedded Systems, Computational Intelligence and Telematics in Control (CESCIT 2021), Jul 2021, Valenciennes, France. pp.141--146, ⟨10.1016/j.ifacol.2021.10.024⟩. ⟨hal-03478234⟩