采用改进非支配近邻免疫算法的低轨混合星座设计优化

针对低轨同构星座覆盖资源在纬度上分布不均匀的不足，提出采用低轨混合星座提升覆盖均匀性的设计方案，并推导了满足全球任意点平均每天覆盖一定次数的最小卫星规模估算公式。针对非支配近邻免疫算法（NNIA）约束处理方面的不足，提出基于约束支配的改进非支配近邻免疫算法（Modified NNIA），并以此设计了一种低轨混合星座优化平台来优化带约束的星座设计问题。仿真结果表明，改进的NNIA算法在收敛速度和多样性上均优于非支配分层遗传算法（NSGA II）和多目标粒子群算法（MOPSO），可大大提高星座设计的效率。同时优化结果也表明低轨混合星座可提高覆盖的均匀性和大部分区域的覆盖次数，进而减少特定覆盖要求所需的卫星数目。

Design Optimization of Hybrid LEO Constellation Using Modified Non Dominated Neighbor Immune Algorithm

As the coverage distribution of homogeneous LEO constellations in latitude is uneven, a hybrid LEO constellation scheme is proposed, so as to enhance coverage uniformity, and a minimum satellite number estimation formula is also deduced to insure a certain number of average coverage times per day at any point of the earth. In the light of the lack of non dominated neighbor immune algorithm (NNIA) in constraint handling, a modified non dominated neighbor immune algorithm (Modified NNIA) based on constrained dominate is suggested, and based on this, a hybrid LEO constellation optimization platform algorithm is designed to optimize the constellation design with constraint. The simulation results show that the proposed Modified NNIA is able to maintain a better convergence speed and also a better diversity feature compared to non dominated sorting genetic algorithm II (NSGA II) and multi objective particle swarm optimization (MOPSO), meanwhile it can improve the efficiency of the constellation design greatly. The optimized results also show that the hybrid LEO constellation can improve the uniformity of coverage and the coverage frequency of most areas, and decrease the number of satellites required for a certain coverage requirements.