基于双延迟深度确定性策略梯度的卫星远程变轨控制  

Satellite remote orbit change control based on twin delayed deep deterministic policy gradient

作　　者：邱鹏鹏 张易诚 曹海涛 郑君铮 Qiu Pengpeng;Zhang Yicheng;Cao Haitao;Zheng Junzheng(School of Computer Science and Technology,Zhejiang Sci-tech University,Hangzhou,Zhejiang,310018,China;School of Information Science and Engineering,Zhejiang Sci-tech University)

机构地区：[1]浙江理工大学计算机科学与技术学院,浙江杭州310018 [2]浙江理工大学信息科学与工程学院

出　　处：《计算机时代》2023年第11期90-93,共4页Computer Era

摘　　要：在间断性点火与小脉冲作用下的卫星,实现椭圆轨道转移极具困难。因此引入深度强化学习双延迟深度确定性策略梯度算法(Twin Delayed Deep Deterministic policy gradient algorithm,TD3),实现了卫星的远程变轨控制。首先建立合理的卫星变轨模型;其次利用TD3算法来模拟卫星点火操作,同时通过设计多种奖励函数引导卫星不断学习,最终到达目标轨道附近;最后通过仿真实验验证了所提TD3算法能够有效控制卫星到达目标轨道附近。It is very difficult to realize elliptic orbit transfer of satellites with intermittent ignition and small pulses.Therefore,Twin Delayed Deep Deterministic policy gradient algorithm(TD3)is introduced to realize remote orbit change control of satellites.Firstly,a reasonable satellite orbit change model is established.Then,the TD3 algorithm is used to simulate the satellite ignition operation,and various reward functions are designed to guide the satellite to keep learning and finally reach the target orbit.The simulation experiments verify that the proposed TD3 algorithm can effectively control the satellite to reach the target orbit.

关 键 词：变轨控制 相对运动 目标轨道 深度强化学习 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]