摘要
船舶航向控制与航行的安全性、可操纵性和经济性密切相关。本文将增强型学习算法与混合智能技术相结合 ,应用于船舶运动航向控制 ,克服了通常混合智能算法的学习需要一定数量样本数据的缺陷 ,又能发挥各种智能算法的优势。仿真结果表明在缺少样本数据情况下 ,该算法可以在一定程度上改进控制效果。
Ship course control is closely related with navigation security, maneuverability and economy. In the paper, reinforcement learning algorithm and hybrid intelligent technique are integrated and applied to ship steering. In this way, the shortcomings of hybrid intelligent learning algorithm, which must be usually provided with some sample data, are overcome, and the advantages of various intelligent algorithms can be fully utilized. It is indicated with simulation that this algorithm can improve control performances in the case of lacking sample data.
出处
《中国航海》
CSCD
北大核心
2001年第2期1-5,共5页
Navigation of China
基金
国家教委博士学科点专项科研基金 ( 9815 10 1)
教育部<高等学校骨干教师资助计划>项目 ( [2 0 0 0 ] 6 5 )
