Fault Tolerant Control for a Hexacopter Based on Optimal Adaptive Control Allocation

A fault tolerant control approach using optimal adaptive control allocation is proposed for a hexacopter with unknown actuator faults. To facilitate adaptive control allocation, concurrent learning method is used to estimate rotor control effectiveness. As the control allocation is formulated as optimization problem, active set method is employed to incorporate the configuration of the priority of the actuator and the pseudo control in the control allocation. One advantage of the proposed approach is its capability of addressing unknown actuator faults online without reconfiguring the controller. Simulation results show that the proposed approach allows the controller to recover to its nominal performance with rotor effectiveness loss.