The invention discloses a failure spacecraft attitude rapid adjustment method based on reinforcement learning, and the method comprises the steps: S1, building a failure spacecraft attitude mathematical model and a constraint model based on spacecraft attitude tail end constraint; s2, based on a Long-term performance index function in a reinforcement learning algorithm, establishing a judgment standard and a Critic network; and S3, on the basis of a Backstepping control framework, an Action network and the Critic network are combined, and an adaptive control method is established so as to control the failed spacecraft to enter a tail end constraint domain. According to the method, rapid attitude adjustment of the failed spacecraft before attitude motion evolution is realized, and a predetermined ignition maneuvering direction is entered.
本发明公开了一种基于强化学习的失效航天器姿态快速调整方法,包括:步骤S1、基于航天器姿态末端约束,建立失效航天器姿态数学模型与约束模型;步骤S2、基于强化学习算法中的Long‑term性能指标函数,建立评判标准和Critic网络;步骤S3、基于Backstepping控制框架结合Action网络和所述Critic网络,建立自适应控制方法,以控制失效航天器进入末端约束域。本发明实现失效航天器在姿态运动演化前的快速姿态调整,进入预定点火机动指向。
Failure spacecraft attitude adjustment method based on reinforcement learning
一种基于强化学习的失效航天器姿态调整方法
2023-04-18
Patent
Elektronische Ressource
Chinesisch
IPC: | B64G Raumfahrt , COSMONAUTICS |
REINFORCEMENT LEARNING FOR SPACECRAFT ATTITUDE CONTROL
TIBKAT | 2020
|Spacecraft attitude preset time cooperative control method based on reinforcement learning
Europäisches Patentamt | 2023
|Spacecraft attitude planning method under complex constraint based on deep reinforcement learning
Europäisches Patentamt | 2023
|