Comparison of three filters in asteroid-based autonomous navigation

Abstract

At present, optical autonomous navigation has become a key technology in deep space exploration programs. Recent studies focus on the problem of orbit determination using autonomous navigation, and the choice of filter is one of the main issues. To prepare for a possible exploration mission to Mars, the primary emphasis of this paper is to evaluate the capability of three filters, the extended Kalman filter (EKF), unscented Kalman filter (UKF) and weighted least-squares (WLS) algorithm, which have different initial states during the cruise phase. One initial state is assumed to have high accuracy with the support of ground tracking when autonomous navigation is operating; for the other state, errors are set to be large without this support. In addition, the method of selecting asteroids that can be used for navigation from known lists of asteroids to form a sequence is also presented in this study. The simulation results show that WLS and UKF should be the first choice for optical autonomous navigation during the cruise phase to Mars.

A mission concept for the low-cost large-scale exploration and characterisation of near earth objects

“…27 shows the distribution of true flyby distances for each asteroid mass, when noise terms are added to the initial condition with standard deviations on position and velocity components of r ¼ 5 km and r ¼ 10 cm/s respectively as order-of-magnitude estimates of the possible uncertainty of the state elements relative to the asteroid from the guidance, navigation and control subsystem. These estimates were obtained by looking at past missions and concepts for deep space navigation relative to asteroids (Ozaki et al, 2016) (Accomazzo et al, 2010) (Hashimoto et al, 2010) (Cui and Zhu, 2014). Fig.…”