基于Advanced-TD3算法的卫星探索控制策略  

Satellite exploration control strategy based on Advanced-TD3 algorithm

作　　者：马志昕 骆淑云 MA Zhixin;LUO Shuyun(of Computer Science and Technology(Artificial Intelligence),Zhejiang Sci-Tech University,Hangzhou)

机构地区：[1]浙江理工大学计算机科学与技术(人工智能)学院,杭州310018

出　　处：《智能计算机与应用》2024年第4期83-88,共6页Intelligent Computer and Applications

摘　　要：卫星控制算法在卫星控制领域拥有十分重要的地位,而深度强化学习则是当前前沿的卫星控制算法之一。针对目前太空环境日渐复杂的问题,提出了基于TD3算法的改进TD3(Advanced-TD3)算法,实现控制卫星到达预定目标区域。在开源环境中进行仿真实验,实验结果验证了该算法的空间探索能力,拥有较高的鲁棒性,可以较为精确地帮助卫星完成控制问题,增强卫星对复杂空间中的控制能力,提高卫星的运行效率。Satellite control algorithms have a critical position in the field of satellite control,Deep reinforcement learning is one of the current cutting-edge satellite control algorithms,In response to the current problem of an increasingly complex space environment,Improved TD3(Advanced-TD3)algorithm a satellite control method based on the improved TD3 algorithm is proposed,which is used to control the satellite to reach the target point automatically.The algorithm is simulated in an open-source environment under Python,and the experimental results verify the space exploration capability of the algorithm,which possesses high robustness and can help the satellite to complete the control problem more accurately in order to enhance the control capability of the satellite in the complex space and improve the operation efficiency of the satellite.

关 键 词：深度强化学习 Advanced-TD3算法 卫星控制 空间探索 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] P185.18[自动化与计算机技术—控制科学与工程]