船舶拟人智能避碰决策理论的集成机器学习策略  

Integrated Machine Learning Strategy of PIDVCA Theory

作　　者：李丽娜[1] 王俊玲[1] 陈国权[1,2] 

机构地区：[1]集美大学,福建厦门361021 [2]大连海事大学,辽宁大连116026

出　　处：《信息与控制》2011年第3期359-368,共10页Information and Control

基　　金：国家自然科学基金资助项目(60774066)

摘　　要：简述了研究船舶拟人智能避碰决策(PIDVCA)的意义、目标及其实现方法,提出了基于机器学习构建动态避碰知识库的关键技术及PIDVCA埋论的机器学习机制,并结合避碰仿真实例,着重讨论了PIDVCA理论的集成机器学习策略以及获取动态避碰知识的机理.The paper summarizes the significance and aim of studying PIDCVA（personifying intelligent decision-making for vessel collision avoidance）,and the way of its realization.As the machine learning is the key technology to construct the dynamic anti-collision knowledge library,the mechanism of machine learning of PIDVCA theory is put forward.The integrated machine learning strategy of the PIDVCA theory and mechanism of acquiring dynamic collision avoidance knowledge are discussed by combination of collision avoidance simulation instance.

关 键 词：PIDVCA 集成机器学习策略 动态避碰知识库 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]