Autonomous estimation of the gravity field for asteroids missions

In the few past decades, there has been an increasing interest in asteroid missions. Many successful asteroid missions have been carried out, leading to many technical advances, particularly in terms of navigation. However, they all faced many challenges, the latter being the Osiris-Rex probe sample collection. To ensure secure navigation around asteroids, they have shown that precise navigation has been and would be of significant importance due to the asteroid irregular gravity field. Through estimating the asteroids’ gravity field, this thesis will approach autonomous navigation. In the navigation software, the spherical harmonics model has been selected to be applied in the simulator. While this approach is convenient and can lead to sufficiently accurate results, it is important to take into account that this model is limited and can not be used for precise landing near the surface, since when approaching the sphere of Brillouin, the spherical harmonics models tend to diverge. Another gravity model must then be used for the last step of the mission. A “real” environment was simulated, taking into account solar induced perturbations and torques as well as the spherical harmonics gravity field of 433 Eros up to degree and order 22. To compare their results, an Extended Kalman filter (EKF) has been developed along with an Unscented Kalman filter (UKF). Although the difference in efficiency was not seen to be significant, the UKF, when calculating large number of parameters, was chosen over the EKF for the simplicity of implementation. To evaluate its performance, the simulator was tested under different conditions, changing the range, inclination of the orbit, and evaluating the effect of different perturbations on the estimation. The simulator was applied to the Near-Shoemaker mission conditions, where it has been shown that all degree and order coefficients up to degree 8 could be estimated with an error below 10%.