Comparative study for Robotic Airship State estimation with LPF, EKF and UKF

This work presents a comparative study between three approaches of pose and velocity estimation for an robotic airship. The first approach is composed of multiples second order Low-pass filters applied to GPS and IMU measured data. The second approach consists of the sensor fusion between GPS, IMU, Barometer and Thermometer with an Extended Kalman Filter (EKF) based in the kinematic equations of motion for a six degrees of freedom (6-DOF) vehicle. Finally, the same 6-DOF equations of motion and sensors are used for the design of an Unscented Kalman Filter (UKF) as a third estimation approach. As result, we obtain greater precision with EKF with a minor advantage over the UKF.