期刊
IEEE ACCESS
卷 7, 期 -, 页码 170374-170384出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/ACCESS.2019.2955440
关键词
Artificial intelligence; Trajectory; Navigation; Reconstruction algorithms; Marine vehicles; Training; Base stations; Automatic identification system data; empirical mode decomposition; Fermat's spiral; trajectory reconstruction; unmanned surface vessel
资金
- Natural Science Foundation of Hubei Province
- Fundamental Research Funds for the Central Universities [2018IVB084, 2017-YB-030, 175111001]
The unmanned surface vessel (USV) trajectory with spatial and temporal information plays an important role in its positioning and navigation. Unlike traditional trajectory reconstruction methods, this paper proposes a novel method based on the automatic identification system (AIS) for USV. Aside from the AIS data applied for restoring the USVs trajectory, the proposed method considers the constraints of the vessels navigation state, maneuvering factors and time stamps. This method consists of three steps: AIS data restoration, Empirical Mode Decomposition (EMD) denoising and Fermats spiral fitting. Firstly, AIS data restoration is applied to eliminate abnormal data. Next, the EMD noise reduction algorithm is used to reduce the jitters and interference of Speed over Ground (SOG) and Course over Ground (COG) components, and the denoised latitude and longitude positions are calculated using the kinematics model of the vessel. Finally, a curve fitting process based on Fermats spiral is employed to reconstruct a smoother trajectory. Several experiments illustrate that the low-cost AIS equipment is compatible with the navigation system of the USV. The reference trajectory is determined by Global Position System (GPS) and Inertial Measurement Unit (IMU) modules in USV. Compared to conventional methods, the residual errors of the proposed method is smaller. The results show that the novel trajectory reconstruction method is effective and can be applied to USVs positioning and navigation.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据