遗传算法在星座多星接近轨道优化中的应用

Genetic Algorithm for Orbital Optimization to Approach Multiple Constellation Satellites

以三颗非共轨的Walker星座卫星为研究对象,对航天器无需变轨与其接近的可能性进行研究.将Lambert方法得到的航天器轨道作为初始轨道,利用遗传算法对初始轨道进行优化.对初始轨道在参考时刻位置和速度的改变量进行编码,形成对应的种群.以航天器与星座卫星之间的最近距离为适应度函数,通过种群的繁殖得到优化结果.结合仿真算例,分析了最小二乘算法和遗传算法在轨道优化中的优劣以及接近过程中轨道摄动的影响.结果表明,遗传算法适用于所提出的轨道改进问题.研究结果可为单航天器无需变轨对星座多星接近问题提供理论依据.

Taking three Walker constellation satellites which locate in different orbit as objects, the probability of a spacecraft approaching them without orbital maneuver is studied. The genetic algorithm is used to optimize the initial orbit which is obtained by Lambert method. The position and velocity＇s variation of the initial orbit on the reference time is encoded to form population. The minimum distance between the spacecraft and three constellation satellites is adopted as fitness function. The optimization result can be attained through the population＇s propagation. Finally, followed by the simulation, the performances of the least square method and the genetic algorithm are analyzed. At the same time, the orbital perturbation in the approaching process is taken into account. The genetic algorithm is suited for this problem of orbital optimization. The study results can be taken as the theoretical proof for a single spacecraft close approaching multiple constellation satellites without orbital maneuver.