Closed-Loop Control of Risley Prism Based on Deep Reinforcement Learning
文献类型：会议论文
| 作者 | Yuxiang, Yao; Ke, Chen; Jinying, Li; Congming, Qin |
| 出版日期 | 2020-03-01 |
| 会议日期 | March 27, 2020 - March 29, 2020 |
| 会议地点 | Guangzhou, China |
| DOI | 10.1109/ICCEA50009.2020.00109 |
| 页码 | 481-488 |
| 英文摘要 | The research of the Risley prism system in the closedloop tracking control of the target has always been focused on the difficulties and hotspots of the system. Due to the nonlinear and strong coupling between the prisms of the system, the currently proposed methods require a large amount of nonlinear and strongly coupled solving. In order to avoid a large number of solving tasks, this paper is based on a control model combining deep reinforcement learning based on Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm and Risley prism control system, combined with the decision-making ability of deep reinforcement learning and the ability of deep neural networks to handle huge state space and continuous action space. The experimental results show that the training of the decision network finally achieves closed-loop control of the target, avoiding the task of solving between prisms, and the model has fast convergence speed and high stability. © 2020 IEEE. |
| 会议录 | Proceedings - 2020 International Conference on Computer Engineering and Application, ICCEA 2020
 |
| 文献子类 | 会议论文 |
| 语种 | 英语 |
| 源URL | [http://ir.ioe.ac.cn/handle/181551/9893]  |
| 专题 | 光电技术研究所_光电工程总体研究室（一室） |
| 作者单位 | Institute of Optics and Electronics, Chinese Academy of Sciences, University of Chinese Academy of Sciences, Chengdu, China |
| 推荐引用方式 GB/T 7714 | Yuxiang, Yao,Ke, Chen,Jinying, Li,et al. Closed-Loop Control of Risley Prism Based on Deep Reinforcement Learning[C]. 见:. Guangzhou, China. March 27, 2020 - March 29, 2020. |
入库方式： OAI收割
来源：光电技术研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。