摘要
为了模拟自主式水下航行器(AUV)在复杂海洋环境中对未知环境的感知以及避障过程,设计了AUV空间避障模拟试验系统.介绍了模拟平台的硬件结构和参数设计,应用激光测距系统(LMS)模拟水下声纳,建立了AUV空间避障仿真环境,设计了基于Q-Learn ing的避障算法,并完成了模拟试验.试验结果表明:AUV空间避障模拟平台可以较真实地反映AUV在水下复杂环境中的空间运动,避障算法可行有效.
To simulate sensation and obstacle avoidance in complex and unknown environment for AUV （autonomous underwater vehicle） , spacial obstacle avoidance simulation test was designed. Hardware architecture and parameters design were introduced, LMS （laser measurement system） was presented to simulate underwater sonar, a simulation environment was constructed and the obstacle avoidance algorithm based on Q-Learning was designed. Finally, simulation tests were carried out. The results show that spacial motion of AUV in complex underwater environment is reflected authentically, and the obstacle avoidance algorithm is practical and effective.
出处
《应用科技》
CAS
2009年第6期53-57,共5页
Applied Science and Technology
关键词
自治式水下机器人
避障
激光测距系统
Q学习
autonomous underwater vehicle
obstacle avoidance
laser measurement systems
Q-Learning
