Growing fast and steady in space: Distributed rapid self-reconfiguration motion planning optimization methods for swarm intelligent space modular self-reconfigurable satellites

The large-scale self-reconfiguration of Space Modular Self-Reconfigurable Satellites (SMSRS) faces critical challenges in the computational complexity of global assignment and motion space problems, low efficiency, as well as the low completion rate associated with the hollow structure, severely limiting their application in extraterrestrial infrastructure construction. In this study, we propose a Rapid Self-Reconfiguration Motion Planning Optimization (RSRMPO) method based on the distributed framework, specifically designed for pivoting cube modular systems. The proposed framework resolves the dual bottlenecks of near-optimal NP-complete problem computational demands and hollow structure formation through a decoupled optimization strategy: assignment optimization at the high-level module ensures structural integrity, while motion planning optimization at the execution layer guarantees swarm fast collision-free coordination. This study contains a Virtual Connection technique, a Degeneration-Growth Assignment (DGA) method, a real-time Feasible Space map (FS_map) generation technique, an energy-cost-optimized A* path planning algorithm, and a collision avoidance technique. In reconfiguration with size spanning 10-1,200 modules, RSRMPO framework demonstrates less planning time and energy consumption and achieves a higher completion rate compared to the Graph-Based Configuration Search (GBCS) algorithm. Simulation experiments confirm that RSRMPO framework provides near-optimal solutions to the NP-complete reconfiguration problem, effectively improving computational efficiency and structural robustness, which establishes significant implications for orbital intelligent large-scale assembly,on-orbit service, and deep-space exploration missions.