Nonlinear cubature Kalman filter for satellite autonomous navigation system with astronomical information

Abstract

All-time and autonomous satellite navigation plays very important roles in the space technology application. In this paper, a state filter and prediction algorithm combined with a cubature Kalman filter (CKF) and a numerical integration method is proposed to realize the all-time and autonomous navigation for the satellite system. The navigation system takes the satellite's on-orbit dynamics model and the astronomical angle information between the sun, the earth and the moon as the state equation and the observation equation, respectively. When the astronomical information is available, the nonlinear CKF algorithm is used to realize the state filter in view of the nonlinearity of the navigation system. Moreover, during the eclipse period, the state prediction algorithm is activated to predict the orbit with the numerical integration method based on the state equation. The algorithm step is established and the performance is compared with unscented Kalman filter (UKF). Finally, simulation results demonstrate the validity and efficiency of the proposed method. The all-time and autonomous navigation precision can be controlled less than 2km, which is enough to satisfy some practical satellite applications.