Crater-based attitude and position estimation for planetary exploration with weighted measurement uncertainty. (November 2020)

• Record Type: Journal Article
• Authors: Zhu, Shengying; Xiu, Yi; Zhang, Ning; Xu, Rui; Cui, Pingyuan
• Is Part Of: Acta astronautica. Volume 176(2020)
• Page Start: 216
• Page End: 232
• Publication Date: 2020-11
• Subjects: Planetary exploration -- Crater-based optical navigation -- Attitude and position estimation -- Measurement uncertainty -- Error ellipse; Astronautics -- Periodicals; Outer space -- Exploration -- Periodicals; Astronautics; Periodicals; 629.405
• DOI: 10.1016/j.actaastro.2020.06.033
• Languages: English
• ISSNs: 0094-5765

Abstract

Abstract: A large number of crater features exist on the surface of interplanetary bodies. Autonomous navigation based on these crater features can obtain excellent navigation performance, which is one of the most important navigation methods for future planetary exploration. This paper presents a new method to estimate the attitude and position of spacecraft based on crater measurement uncertainty. Firstly, the error distribution of craters' localization is introduced, considering the characteristics of edge detection in crater images. Then, the error uncertainty of crater localization is described by the error ellipse and the influence of related factors on the crater localization error is analyzed. Further, in consideration of the characteristics that the localization errors are anisotropic, correlated and non-identically distributed, the weighted matrix of different craters is constructed by singular value decomposition (SVD) of the error uncertainty matrix. Thereafter the weighted matrix is integrated into the attitude and position estimation algorithm. As a result, the weighted measurement uncertainty method for crater-based pose estimation is formed. Finally, the proposed algorithm is verified by Monte Carlo simulation. Highlights: A weighted method with uncertainty for crater-based pose estimation is proposed. The expression of measurement uncertainty is derived based on optical information. Measurement uncertainty is anisotropic, correlated and non-identically distributed. The uncertainty can be decomposed to construct weighted matrix for estimation.