• stepwise regression
• system identification
• angle of sideslip
• angle of attack
• virtual sensor

The aim of this thesis is to design Virtual Sensors to estimate the angle of attack and sideslip. A Virtual Sensor can be defined as a software module which utilizes measurable signals (aircraft accelerations, in the present work) in order to reconstruct a signal of interest (aerodynamic angles). The use of Virtual Sensors can improve the reliability of the Air Data System, introducing an analytical redundancy and increasing the fault detection performance. Actually, Virtual Sensors are interesting as they may be used to replace physical sensors for fault diagnosis, reducing the redundancy of on-board hardware and thus aircraft cost and weight. The algorithm developed for the estimation of both angle of attack and angle of sideslip is based on flight data from Aermacchi M-346 MASTER. The key idea behind this estimation is to exploit the relation between aerodynamic angles, aerodynamic forces acting on the aircraft and accelerations. As accelerations are measurable and aerodynamic forces can be modeled, it is possible to estimate the aerodynamic angles.