Abstract: During a safety and accurately celestial body pinpoint landing, the spacecraft must realtimely and precisely determine its relative position and velocity to the body. To overcome the disadvantage that the slant range and the relative velocity measurements can’t well revise the horizontal position estimation obtained by inertia navigation, and to fuse multisource information, an autonomous navigation method for celestial body pinpoint landing is proposed based on landmarks image and covariance intersection algorithm. In this method, an optical camera is added to the integrated navigation system, and takes pictures of landmarks to obtain the essential horizontal position information. To esitmate the states, a distributed extended Kalman filter based on the covariance intersection is used to fuse the multisource information optimally, which can improve the precision and the robustness. Numerical simulations demonstrate the effectiveness of the proposed approach.

Key words: pinpoint landing, landmark image, information fusion, autonomous navigation