Nonlinear estimation for spacecraft attitude using decentralized unscented information filter

This paper presents a spacecraft attitude estimation algorithm using the unscented information filter. The objective of the proposed algorithm is to make the spacecraft attitude accurately estimate with low computational load. The main contribution of this study is to derive the unscented information filter for nonlinear spacecraft attitude dynamics and multiple sensor systems. For the multiple sensor estimation, the decentralized unscented information filter is used in this study. This filter provides accurate attitude estimation from the unscented filtering and less computational time from the information filtering. To verify the effectiveness of the proposed decentralized unscented information filter algorithm, numerical simulations are performed for a spacecraft system with three gyroscopes.