Active fault tolerant control of an octorotor UAV

Redundancy is the key point to design powerful fault tolerant and reliable controllers for dynamic systems. In this chapter, Dynamic Control Allocation method is used to develop an active Fault Tolerant Control (FTC) scheme for the Octorotor redundant UAV system. A Sliding Mode Observer is used for Fault Detection and Isolation (FDI) in order to detect, online, any fault or failure affecting the UAV motors during flight. Whenever a severe fault or a total failure of a rotor or a group of rotors are detected and measured, the gain vector of the dynamic control allocator is updated. This update results in the redistribution of the control effort on healthy actuators, which reduces the effect of the faulty actuators on the overall control vector. The main controller of the octorotor is based on Sliding Mode Control (SMC) technique which is a robust nonlinear control method with considerable immunity against disturbances and uncertainties. The proposed method is compared with Passive Fault Tolerant Sliding Mode Control, Adaptive Active Fault Tolerant Control SMC, and with a similar technique using Pseudo-Inverse method for Control Allocation. Simulation results using Matlab/SimulinkTM show that Fault Tolerant Controllers based on Control Allocation methods outperform both Passive and Active Fault Tolerant Controllers. Dynamic Control Allocation was able to compensate for the effect of multiple severe faults while. taking into account actuator and time constraints. On the other hand, using Pseudo-Inverse Control Allocation method is straightforward and effective, but it reduces the applicability and capabilities of the controller. Here, the actuators speeds and their position limits are not taken into account and the control allocation depends only on the current control demands. Three fault types were used to evaluate the controllers, the loss of effectiveness faults, stuck faults, and additive faults.