Cooperative Navigation and Landmark Set Construction Method for Multi-Spacecraft Orbital Exploration of Small Bodies

In small body orbital exploration, long communication delays necessitate high spacecraft autonomy. Optical cameras, offering rich environmental perception, are essential low-cost sensors for autonomous navigation. Multi-spacecraft cooperative exploration further enhances observation coverage and system robustness, making cooperative navigation a key research focus. This paper proposes an optical camera-based method for cooperative navigation and landmark set construction method for multi-spacecraft orbital missions. During initialization, each spacecraft integrates multi-frame visual odometry with initial orbit and attitude data to estimate its state and construct a landmark set in the small body's body-fixed frame. During operation, an adaptive density-based clustering algorithm (DBSCAN) dynamically extracts dense and stable core landmarks. A weighted fusion and Iterative Closest Point (ICP) registration strategy then aligns and integrates landmark sets across spacecraft into a fused surface landmark set. This fused landmark set supports ongoing state propagation and landmark updating, enabling continuous optimization and collaborative mapping. Simulations verify that the proposed method improves both mapping accuracy and state estimation robustness, supporting autonomous navigation for deep-space multi-spacecraft missions.