New integrated attitude and orbit filter for non-cooperated satellite formation

To enhance the attitude perception of the satellites and engineering redundancy in the space non-cooperative mission, a filtering algorithm based on the integration of attitude and relative orbit while without acknowledging angle information was presented which combined with the attitude and the coupling information of the two satellites' relative position. The error quaternion-based attitude equation under inertial frame and the relative motion equation under satellite orbit frame were used as the filter's system equation, and the joint efficient estimation of satellite attitude stability, zero bias of gyros and relative position and velocity was realized by using extended Kalman filtering (EKF). Comparison and analysis were made on two classical satellite formation models (leader-follower and fly-around) and the filter's local observability in different structure and scale in relative motion. The effectiveness of the algorithm is validated by some simulation experiments. Simulation results show that the relative position estimation error and velocity estimation error were 0.3 m and 10^-4 m/s respectively, the attitude determination error was 0.05° and the gyro bias estimation error was 10^-4 (°)/s, which satisfy the mission requirements for satellite formation.