Particle filter based integrated navigation for quad rotors

摘要

In this work, in order to achieve increased positioning accuracy, a particle filter based integrated navigation system is designed for a rotary wings unmanned aerial vehicle (UAV). As the navigation problem has highly non-linear characteristics on the system and measurement models, the designed algorithm which is based on particle filter works more effectively with respect to Kalman filter-based algorithms which have strict linearity assumptions. In order to verify the proposed algorithm, firstly the designed navigation system is tested with a flight scenario planned in an only simulation environment. After that, the designed algorithm is tested with the data acquired on a real rotary wings system in the laboratory for a different number of particles separately and it has been shown to produce satisfactory results. In this study, an integrated navigation solution having high accuracy and reliability was calculated for a quadrotor having six degrees of freedom.