| 基于增强式学习的仿生机器鱼避障控制 |
| |
| 引用本文： | 沈志忠,曹志强,谭民,王硕.基于增强式学习的仿生机器鱼避障控制[J].高技术通讯,2006,16(12):1253-1258. |
| |
| 作者姓名： | 沈志忠 曹志强 谭民 王硕 |
| |
| 作者单位： | 1. 北京科技大学信息工程学院,北京,100083;中国科学院自动化研究所复杂系统与智能科学重点实验室,北京,100080
2. 中国科学院自动化研究所复杂系统与智能科学重点实验室,北京,100080 |
| |
| 基金项目： | 国家高技术研究发展计划(863计划)
,
国家自然科学基金 |
| |
| 摘 要： | 研究了仿生机器鱼的自主避障控制,提出了一种集成超声和红外传感器的仿生机器鱼系统设计方案,为机器鱼安装了3个超声传感器和3个红外传感器.针对此仿生机器鱼系统,提出了一种基于增强式学习的仿生机器鱼自主避障控制策略,给出了状态集和行为集,采用当场奖励和延时奖励相结合的方法,通过学习获得了有效的状态行为组合.仿真实验验证了学习结果的有效性.
|
| 关 键 词： | 仿生机器鱼 避障 学习 |
| 收稿时间： | 2005-11-18 |
| 修稿时间： | 2005年11月18 |
Control of obstacle avoidance of biomimetic robot fish based on reinforcement learning |
| |
| Shen Zhizhong,Cao Zhiqiang,Tan Min,Wang Shuo.Control of obstacle avoidance of biomimetic robot fish based on reinforcement learning[J].High Technology Letters,2006,16(12):1253-1258. |
| |
| Authors: | Shen Zhizhong Cao Zhiqiang Tan Min Wang Shuo |
| |
| Affiliation: | 1. School of Information Engineering, University of Science and Technology Beijing, Beijing 100083; 2. Laboratory of Complex Systems and Intelligence Science, Institute of Automation, Chinese Academy of Sciences, Beijing 100080 |
| |
| Abstract: | Autonomous control of obstacle avoidance of biomimetic robot fish is studied. The design of a kind of biomimetic robot fish system with ultrasonic and infrared sensors is given. Three ultrasonic sensors and three infrared sensors are equipped with the robot fish. Aiming at this robot fish system, an obstacle avoidance strategy based on reinforcement learning is proposed. The state and behavior sets are brought forward. Through adopting immediate and delayed rewards, valid state-behavior pairs are obtained through learning. Computer simulations show the validity of the learning results. |
| |
| Keywords: | biomimetic robot fish obstacle avoidance learning |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
