基于微分进化算法的深空轨道全局优化设计
THE GLOBAL OPTIMIZATION OF SPACE EXPLORATION TRAJECTORY DESIGN BASED ON DIFFERENTIAL EVOLUTION ALGORITHM
-
摘要: 针对多目标多任务的深空探测轨道设计问题,提出一种新的将探测目标、探测方式、探测顺序以及发射窗口同时作为优化变量, 并采用微分进化算法进行全局优化的设计方法. 使用该方法在只考虑太阳中心引力作用的二体模型下,基于圆锥曲线拼接法建立第三届全国深空轨道设计竞赛问题的优化模型并进行求解. 最后利用该方法求解ESA的ACT研究团队的深空探测任务算例并对结果进行对比分析. 结果表明, 提出的全局优化设计方法对解决多目标、多任务深空探测轨道优化设计问题是可行和有效的.Abstract: In this paper, a new global optimization method is proposed to solve the multi-target space exploration trajectory design problem. The variables of targets selection, visiting sequences and launch windows are synchronously optimized. In the context of two-body model in which the central gravity of the sun is only considered with the thrust as impulse, the patched-conic method accompanied with differential evolution (DE) algorithm are applied to solve the global optimization trajectory design problem. Then the method is utilized to solve the problem of the Third National Design Competition for Deep Space Exploration and the example missions of the ACT of ESA. It proves that the global trajectory optimization design method is feasible and valuable for the problem of multi-target and multi-mission space trajectory design.