Application of extended Kalman filter towards UAV identification

Abstract

This chapter considers the application of Extended Kalman Filtering (EKF) towards Unmanned Aerial Vehicle (UAV) Identification. A 3 Degree-Of-Freedom (DOF), coupled, attitude dynamics is considered for formulating the mathematical model towards numerical simulation as well as Hardware-In-Loop (HIL) tests. Results are compared with those obtained from a state-space estimation method known as the Error Mapping Identification (EMId) in the context of processing power and computational load. Furthermore the statistical health of the EKF during the estimation process is analysed in terms of covariance propagation and rejection of anomalous sensor data. © 2007 Springer-Verlag Berlin Heidelberg.