Fault-Tolerant Control and Fault Detection for Unmanned Aerial Vehicles

The purpose of this diploma thesis is the presentation of some proposals for the fault tolerant control of unmanned aerial vehicles (UAV). In the first part, the model reference adaptive control (MRAC) scheme is utilized to control UAVs susceptible to actuator failures. In the second part, the multiple model adaptive estimation method (MMAE) is used for the detection and isolation of either actuator or sensor failures. In order to widen the class of detectable failures, the MMAE method is upgraded with extended Kalman filters (EKF) and thus becomes the extended multiple model adaptive estimation method (EMMAE). In this context, each EKF is used for state vector estimation on the one hand and for the estimation of