摘要
针对船舶AIS数据丢失或错误等问题,本文借助分段三次Hermite插值实现AIS数据初步修复或预测,建立神经网络训练集和测试集,开展单点和连续多点AIS数据修复和预测;对比分析了BP神经网络与三次样条插值、分段三次Hermite插值方法以及组合算法在船舶AIS数据修复和预测中的精度。以重庆弯曲河段和武汉顺直河段为例,分析了航道平面形态、算法组合等对于船舶AIS数据修复和预测精度的影响。结果表明:联合算法有效提升了船舶AIS数据修复精度;在船舶AIS预测中,神经网络模型表现最优。研究成果可为船舶行为特征分析、建模等相关领域的研究提供借鉴。
We established models for single-point and continuous multipoint automatic identification system (AIS) data restoration and prediction on the basis of the training and testing sets of the BP neural network and the piecewise cubic Hermite interpolation to resolve AIS data loss or error. The precisions of the BP neural network, traditional cubic spline interpolation,piecewise cubic Hermite interpolation and combinecl algorithm in ship AIS data restoration and prediction were compared. Finally,the Chongqing and Wuhan Waterways were investigated in case studies. The influence of river configuration and algorithm combinations on the accuracy of AIS data restoration and prediction models was analyzed. Results indicated that the data repairing accuracy of the combined algorithm is dramatically improved,and the neural network model demonstrated the best AIS data prediction performance. Results can provide references to research on ship behavior characteristic analysis,modeling,and other related fields.
作者
刘磊
蒋仲廉
初秀民
钟诚
张代勇
LIU Lei;JIANG Zhonglian;CHU Xiumin;ZHONG Cheng;ZHANG Daiyong(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;School of Logistics Engineering,Wuhan University of Technology,Wuhan 430063,China)
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2019年第6期1072-1077,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金项目(51479155,51709220)
中央高校基本科研业务经费专项资金(2017-zy-0479)
关键词
水路运输
数据修复和预测
BP神经网络
分段三次Hermite插值
三次样条插值
联合数学模型
自动识别系统数据
修复和预测精度
waterway transportation
data restoration and prediction
BPneural network
piecewise cubic Hermite interpolation
cubic splineinterpolation
joint mathematical model
automatic identification system(AIS) data
repair and prediction accuracy
