Autonomous station-keeping strategy for geostationary satellite with electric propulsion system

Geostationary orbit (GEO) satellite guidance and station keeping often use mean elements as inputs. When it comes to geostationary orbit with small eccentricity and inclination, traditional estimation method of the mean orbital elements is not robust enough to meet the required precision, which results from a total difference between high orbital perturbations and those in low orbit. The purpose of this paper is to develop an efficient method for autonomous onboard estimation of the mean orbital elements of GEO satellites. First of all, the dynamical model considering the Earth's non-spherical perturbation, luni-solar perturbation, and solar radiation perturbation is established. According to the semianalytical theory, the effect of the significant perturbation can be analytically evaluated. Then, the mapping from mean to osculating elements is derived. Regarding mean elements as state parameters and osculating elements as estimators, the estimation can be performed by Extended Kalman Filter (EKF). The accuracy and efficiency of the method can be validated through numerical simulations of the station keeping mission.