Adaptive filtering-based relative navigation for formation-flying satellites

This research introduces a novel approach to autonomous relative navigation for dual satellite formations, employing a strategy that utilizes Probabilistic Adaptive Extended Kalman Filters (p-AEKFs) based on Global Navigation Satellite System (GNSS) measurements. The presented p-AEKF algorithm is based on tracking normalized innovation sequences in the filter and calculating the probability of normal operation of the estimation system. The filter gain is adjusted based on this probability to maintain the filter's tracking performance despite faulty measurements. The positions of the mother and follower satellites are estimated using the proposed probabilistic AEKF using GNSS distance measurements. This approach integrates the state vectors generated by the two p-AEKFs using the Hill-Clohessy-Wiltshire (HCW) relative motion model to enhance the accuracy of relative state estimation between the satellites. To estimate the parameters of the HCW relative motion model, it is proposed to use the traditional EKF. Proposed probabilistic and scaling approaches are compared using simulations.