中国空间科学技术 ›› 2025, Vol. 45 ›› Issue (5): 110-120.doi: 10.16708/j.cnki.1000-758X.2025.0079

• 论文 • 上一篇    下一篇

飞跃探测全过程最优轨迹设计及实时控制方法

于萍*,王浩帆,王泽国,付仁皓   

  1. 北京控制工程研究所,北京100190
  • 收稿日期:2024-07-29 修回日期:2025-02-03 录用日期:2025-02-10 发布日期:2025-09-17 出版日期:2025-10-01

Whole-process optimal trajectory design and real-time control for hopping detection mission

YU Ping*,WANG Haofan,WANG Zeguo,FU Renhao   

  1. Beijing Institute of Control Engineering, Beijing 100090,China
  • Received:2024-07-29 Revision received:2025-02-03 Accepted:2025-02-10 Online:2025-09-17 Published:2025-10-01

摘要: 飞跃探测是一种新颖的地外天体探测任务形式,是上升任务和着陆任务的复合过程。研究了兼具飞跃全程轨迹优化及实时性的控制方法。在标称轨迹优化方面,不同于传统的分段优化方法,从飞跃全程优化的视角出发,提出了解析与偏微分方法结合的全程轨迹设计方法。在实时轨迹控制方面,基于预测校正的思想,应用Jacobian矩阵,通过综合修正轨迹参数和推力方向,来确保终端状态的实现。实时轨迹控制中应用的Jacobian矩阵的计算公式均为显示解析形式,且仿真表明仅需2次迭代即可满足控制需求,说明了算法的实时性。在相同的仿真条件下,与离线的全程凸优化算法相比燃料仅相差20g,说明了其燃料优化性能与全程凸优化算法相当。该方法可以适用于常推力飞跃全程轨迹优化及实时轨迹控制,能够保证飞跃任务轨迹的全程最优性和控制的实时性。


关键词: 飞跃转移, 轨迹设计, 解析控制, 深空探测器

Abstract: The hopping detection is a novel form of extraterrestrial body exploration mission, which is a combination of ascent missions and landing missions. The study of control methods that combine the entire process of hopping trajectory optimization and real-time performance is of great significance. In terms of nominal trajectory optimization, unlike traditional segmented optimization methods, a trajectory design method that combines analytical and partial differential approaches is proposed from the perspective of optimizing the entire hopping process. In terms of real-time trajectory control, based on the prediction-correction method, the Jacobian matrix is applied to ensure the achievement of terminal conditions by comprehensively correcting trajectory parameters and thrust directions. The calculation formulas for the Jacobian matrix used in real-time trajectory control are all explicit and analytical, and simulations show that only two iterations are enough to meet control requirements, proving that the algorithm proposed has high real-time performance. With the same simulation conditions, compared with the offline whole-process convex optimization results, the consumed fuel difference between the two methods is only 20g, indicating that its fuel optimization performance is relatively similar to the convex optimization one. This method can be applied to the whole-process thrust hopping trajectory optimization with constant thrust and real-time trajectory control, ensuring the whole-process optimality and real-time control performance.

Key words: hopping detection;trajectory design, analytical control, deep space probe