分布估计算法在航天器近距离最优交会中的应用

针对空间在轨服务飞行器实施近距离变轨最优化问题,探讨了在能量和时间两项指标情况下,航天器轨道机动中能量时间综合最优化的研究方法。基于C-W方程,推导了在轨服务器在双冲量变轨时的特征速度,以时间-燃料为指标构造了相应的模型,并针对基本遗传算法局部搜索能力不强的问题,采用一种新的利用统计学习手段从群体宏观角度建立描述解分布的分布估计算法。仿真结果表明,该分布估计算法可加速算法的收敛,具有良好的优化能力,能够从宏观上对整个群体建立模型,得到了混合指标下时间和能量关系,实现近距离变轨最优指标的精确数值模拟。从数值结果的对比分析中得出了一些有意义的结论,可供下一步研究参考。

Application of the EDA Algorithm in the Near Range Optimal Rendezvous for Spacecrafts

The methods for the optimization of time-energy in the orbit maneuvering of spacecraft under two indexes,with emphasis on the optimization of near range orbit,are explored.By means of the C-W equations,the characteristic velocity is derived for the on-orbit servicing spacecraft using a two-impulse orbit.The numerical calculation shows that EDA can accelerate the convergence of the algorithm,and has good optimal capability.It can also build a macro model for the entire group,getting time and energy relations under the mixed indicators.As a result,an accurate numerical simulation of the optimal indexes in the near range is realized.From the numerical results and comparison between these results,some conclusions of significance are drawn,which provides reference for future research.