基于案例推理的舰船管路破损应急决策模型研究  

Research on Emergency Decision-making Model of Ship Pipeline Damage Based on Case-based Reasoning

作　　者：朱玲娜 任凯[1] 浦金云[1] ZHU Lingna;REN Kai;PU Jinyun(College of Power Engineering,Naval University of Engineering,Wuhan 430033)

机构地区：[1]海军工程大学动力工程学院,武汉430033

出　　处：《舰船电子工程》2022年第9期141-145,150,共6页Ship Electronic Engineering

基　　金：装备预先研究基金项目“损管信息联动管制技术研究”(编号:1020201021001)资助。

摘　　要：舰船存在的内部结构复杂并且作业环境严峻,一旦在航行或是战斗中发生破损故障,将可能导致重大灾难的发生。针对此问题,为能快速有效实现舰船的管路损管决策,论文基于传统的案例推理模型,结合BP神经网络与学习型伪度量(LPM)设计了一种舰船管路检索方法。在此基础上形成舰船损管辅助决策技术,帮助舰船指挥员进行损管决策。最后以舰船管路破损为例,将论文的研究结果与其他类型的算法进行比较,验证研究结果的快速性和正确率。The complex internal structure and severe operating environment of ships may lead to a major disaster in case of breakage and failure during navigation or combat.To address this problem,this paper designs a ship’s pipeline retrieval method based on the traditional case-based reasoning model,combined with BP neural network and learning pseudo-metric(LPM),in order to quickly and effectively realize the ship’s pipeline loss and management decision.Based on this,a ship’s pipeline retrieval technique is developed to help ship’s commanders make pipeline loss decisions.Finally,the research results of this paper are compared with other types of algorithms to verify the rapidity and correctness of the research results,taking ship’s pipeline breakage as an example.

关 键 词：应急决策 案例推理 伪度量 BP神经网络 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]