Control Allocation for Multirotor Subjected to a Complete Actuator Failure

In this paper, a new method of control allocation is presented in the context of redundantly-actuated multirotor unmanned aerial vehicles (UAV). The equation between the virtual control efforts and the output of the actuator and the output limit of the actuator are taken as the constraints to express the method as a non-linear convex optimization problem. The proposed method is applied to extreme cases including poor actuator performance and propeller damage of multirotor and is suitable for redistributing the control effort to actuators while respecting its mechanical constraints.