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

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．

Abstract: 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.