Robust Adaptive Kalman Filter for estimation of UAV dynamics in the presence of sensor/actuator faults

In this paper a Robust Adaptive Kalman Filter (RAKF) is introduced. The RAKF incorporates measurement and process noise covariance adaptation procedures (R and Q adaptation respectively) and utilizes adaptive factors in order to adapt itself against sensor/actuator faults. Thus the filter stands robust against the faults and even in case of sensor/actuator failure keeps providing accurate estimation results. In a single algorithm, the RAKF detects the fault, isolates it and applies the required adaptation process such that the estimation characteristic is not deteriorated. The performance of the proposed RAKF is investigated by simulations for the state estimation procedure of an Unmanned Aerial Vehicle.