Extended Kalman filters for close-range navigation to noncooperative targets.

• A set of multiplicative extended Kalman filters for uncooperative relative navigation is developed. • Unit quaternions are used to represent the coupled roto-translational relative dynamics. • The filters estimate the 6 DoF state and – in some options - the main parameters of the Target object. • These filters can serve several architectures of the relative navigation system. • Numerical simulations representative of close-range Target inspection are performed. This work presents a set of dynamic filters for estimating the relative roto-translational state and the main parameters of a noncooperative target from an observing chaser satellite during close proximity operations. The proposed different options address a wide range of design possibilities for the architecture of the relative navigation system. All filters are derived from a common, general, core shaped as a dynamic multiplicative extended Kalman filter using dual quaternions. This allows exploiting the advantages of handling the pose (i.e. , attitude and position) in a multiplicative fashion, while improving the accuracy in the estimation of the angular and linear relative velocities, as well as enabling the estimation of some meaningful parameters of the target spacecraft (e.g. , the ratios of the moments of inertia, position and orientation of the principal axes frame). Moreover, by adopting relative kinematics and dynamics equations in dual quaternions, the inherent coupling of the six degrees-of-freedom motion is addressed with no approximations. All filters take as observations only the noisy pose measurements from an electro-optical device. For each proposed formulation, numerical simulations are carried out to show the behavior of the filter within a scenario representative of close-range target inspection at conclusion of the mid-range rendezvous. [ABSTRACT FROM AUTHOR]