Extended Kalman filter based data fusion algorithm implemented in a Sense and Avoid System for a mini-UAV application

The present article details the implementation of an Extended Kalman Filter as a data fusion strategy for the development of a Sense and Avoid system to be deployed onboard a mini-VAV (maximum weight < 25 kgf). After a brief overview of the overall sense and avoid system, the mathematical formulation of the data fusion strategy is discussed, with particular focus to the tracking algorithm based on the Extended Kalman Filter developed for the specific application. An extensive hardware-in-the-loop simulation campaign, including the artificial generation of realistic optical synthetic imagery and an accurate dynamic model of the aircraft demonstrates the effectiveness of the filter both in rejecting the sensor noise and in coping with different refresh rates of the sensors. Simulation results prove that the suggested algorithm allows for successful tracking of the trajectory of non-cooperative intruder aircraft on a collision course and allows for the generation of an effective evasive maneuver which complies both with aero mechanical constraints of the VAV and with minimum separation distance requirements