Multi-Objective Optimization for Mission-Oriented Earth–Moon Navigation Constellations

With the continuous advancement of cislunar space exploration, reliable navigation infrastructure for diverse users is essential for future deep-space missions. This study investigates the design and service performance of navigation satellite constellations for multiple user types in cislunar space. Considering period coordination among different orbital families, two candidate orbit clusters with different reference periods are constructed using a four-level period structure with a ±1% tolerance. A hierarchical user scenario is established, including transfer-orbit, spatial sampling, and representative validation users. A spatial demand-aware multi-objective evolutionary algorithm (SD-MOEA) is proposed to generate compact CRTBP-based constellation candidates. Two representative constellations are selected from the resulting Pareto-optimal solutions for performance evaluation. The results show that both constellations exhibit geometric service potential under the CRTBP evaluation model. The proportions of time with at least 20 visible satellites for transfer-orbit users reach 97.32% and 75.84%, respectively. For spatial sampling users, Constellation A achieves a maximum visible satellite number of at least 15, while Constellation B provides enhanced service in the near-Moon region. NRHO and lunar-surface evaluations further characterize visibility and geometric availability. The proposed framework supports demand-driven cislunar navigation constellation design and further architecture studies.