强化学习算法应用于船舶运动的混合智能控制

HYBRID INTELLIGENT CONTROL FOR SHIP MOTION BASED ON REINFORCEMENT LEARNING ALGORITHM

将强化学习算法与混合智能技术相结合 ,应用于船舶运动控制 ,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 .

Reinforcement learning algorithm and hybrid intelligent technique are integrated and applied to ship motion control. This paper deals with the defect of usual hybrid intelligent algorithm learning that must be provided with some sample data and the advantages of all kinds of intelligent algorithm are fully utilized. Simulations demonstrate that the algorithm can partly improve control performances if there is a lack of sample data.