首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于BP-DS证据理论的航标状态识别模型研究
引用本文:聂梓熠,蒋仲廉,初秀民,刘怀汉.基于BP-DS证据理论的航标状态识别模型研究[J].水运工程,2022(1):150-156.
作者姓名:聂梓熠  蒋仲廉  初秀民  刘怀汉
作者单位:武汉理工大学;武汉理工大学能源与动力工程学院;长江航道局
基金项目:国家重点研发计划项目(2018YFB1600400);国家自然科学基金项目(52071250、51709220)
摘    要:内河航标标示了可航行水域的边界,是保障内河船舶安全航行的重要前提.面向内河航标不同故障模式,构建一种基于BP-DS证据理论的航标状态识别模型.采用两个BP神经网络分别训练航标基础属性数据与环境因素数据,计算独立的证据值;通过DS证据理论实现航标运行状态识别结果的决策融合.结果表明,基于BP-DS证据理论的航标状态识别模...

关 键 词:航标状态  BP(反向传播)神经网络  DS(Dempster-Shafer)证据理论  决策融合

Research on status recognition model of aids to navigation based on BP-DS evidence theory
NIE Zi-yi,JIANG Zhong-lian,CHU Xiu-min,LIU Huai-han.Research on status recognition model of aids to navigation based on BP-DS evidence theory[J].Port & Waterway Engineering,2022(1):150-156.
Authors:NIE Zi-yi  JIANG Zhong-lian  CHU Xiu-min  LIU Huai-han
Institution:(National Engineering Research Center for Water Transport Safety,Wuhan University of Technology,Wuhan 430063,China;School of Energy and Power Engineering,Wuhan University of Technology,Wuhan 430063,China;Changjiang Waterway Bureau,Wuhan 430010,China)
Abstract:Inland aids to navigation mark the boundaries of navigable waters and play a significant role for ensuring the safe navigation of inland ships.Considering different failure modes of aids to navigation,we propose a status recognition model based on BP-DS evidence theory,use two BP neural networks to train the basic attribute data and environmental factor data,calculate the independent evidence values,and further apply the DS evidence theory to realize the decision-making fusion of aids to navigation operation state recognition results.The results show that the accuracy of status recognition model of aids to navigation based on BP-DS evidence theory is more than 80%,which is higher than the traditional empirical threshold judgment method.
Keywords:status of aids to navigation  back propagation neural network  Dempster-Shafer evidence theory  decision fusion
本文献已被 维普 等数据库收录!
点击此处可从《水运工程》浏览原始摘要信息
点击此处可从《水运工程》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号