首页 >

途经点约束下模型预测与反演制导的遗传算法寻优(航天器智能控制专栏)

谭天乐,史树峰   

  1. 上海航天控制技术研究所
  • 收稿日期:2025-04-22 修回日期:2025-07-10 出版日期:2025-07-15 发布日期:2025-07-15
  • 通讯作者: 谭天乐

Genetic Algorithm Optimization of model predictive and inversive guidance control with waypoint constraints

  • Received:2025-04-22 Revised:2025-07-10 Online:2025-07-15 Published:2025-07-15
  • Contact: Tian-Le TAN

摘要: 面向空间交会对接、编队构型、在轨服务、穿行规避等任务场景需求,在起止状态明确,需经过若干期望途经点的空间多脉冲相对制导过程中,为获得航天器的最优燃料消耗,需优化制导轨迹及控制过程,确定合理的途经顺序、途经时刻和途经速度。在离散时间系统中,利用系统动力学状态转移模型,通过变量替换和递进迭代方法,用已知的状态和待求的制导控制序列表示未知的途经点相对速度,结合广义逆最小二乘解,得到全局二范数最优的控制序列,进而设计出了时间和途经点位置约束下,空间相对制导的模型预测与反演制导控制方法。为进一步优化制导过程燃料消耗,采用遗传算法搜索经过各途经点的途经顺序和最优时刻。近圆轨道交会制导控制的仿真结果表明,在约束/不约束途经点途经顺序、经过时间条件下,所提出的方法能够在与途经时刻、途经顺序相关的三种场景下,实现精度高、代价小、柔滑平顺的的相对制导控制燃料消耗寻优。

关键词: 途经点, 多脉冲制导, 状态转移, 交会对接, 模型预测与反演控制, 离散时间系统, 遗传算法

Abstract: To meet the requirements of mission scenarios such as space rendezvous and docking, formation configuration, in orbit services, and maneuvering avoidance, in the process of space multi pulse relative guidance with clear starting and ending states and passing through several expected waypoints, in order to obtain the optimal fuel consumption of the spacecraft, it is necessary to optimize the guidance trajectory and control process, determine a reasonable passing sequence, passing time, and passing speed.In discrete-time systems, based on the system dynamics state transition model, variable substitution and progressive iteration methods are used to represent the unknown relative velocity of the waypoint using known states and the guidance control sequence to be solved. Combined with the generalized inverse least squares solution, the optimal control sequence under the global two norm condition is obtained. Then, a model prediction and inversion guidance control method for spatial relative guidance under time and waypoint position constraints is designed.To further optimize the fuel consumption during the guidance process, a genetic algorithm is used to search for the sequence and optimal time of passing through each waypoint. The simulation results of near circular orbit rendezvous guidance control show that the proposed method can achieve fuel consumption optimization in high-precision, smooth relative guidance control processes under the conditions of constrained/unconstrained waypoint sequence and passage time in three scenarios related to passage time and sequence..

Key words: waypoint, multi pulse guidance, state transition, rendezvous and docking, model predictive and inversive control, discrete time systems, genetic algorithm

中图分类号: