全文下载:
202601020.pdf
文章编号: 1672-6987(2026)01-0149-10 DOI: 10.16351/j.1672-6987.2026.01.020
张瑞, 郭瑛*, 刘鹏(青岛科技大学 信息科学技术学院, 山东 青岛 266061)
摘要: 水下传感器网络是实现海洋环境监测智能化的重要手段,受海水物理特性的影响,水声信号的传播速率是非线性且不稳定的,这使得基于水声信号的测距方法不可避免地出现较大的测距误差,为降低测距误差对水声无线传感器网络定位精度的影响,本文提出了一种基于卷积长短期记忆神经网络(ConvLSTM)修正测距的水下定位算法,利用基于ConvLSTM构建的神经网络对海洋环境的历史数据进行特征提取,对水声信号速率进行分层次预测,从而达到减小测距误差的目的。将修正过的测距值代入基于移动信标和已定位节点的迭代定位算法中,可以进一步提高水下未知节点的定位精度。仿真实验结果表明:该算法可以显著降低定位误差,具有较高的定位精度与可行性。
关键词: 水下定位算法; 声速预测; ConvLSTM; 水下传感器网络
中图分类号: TP 393 文献标志码: A
引用格式: 张瑞, 郭瑛, 刘鹏. 基于ConvLSTM修正测距的水下节点定位算法[J]. 青岛科技大学学报(自然科学版), 2026, 47(1): 149-158.
ZHANG Rui, GUO Ying, LIU Peng. Underwater node localization algorithm based on convlstm modified ranging[J]. Journal of Qingdao University of Science and Technology(Natural Science Edition), 2026, 47(1): -.
Underwater Node Localization Algorithm Based on ConvLSTM Modified Ranging
ZHANG Rui, GUO Ying, LIU Peng(College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China)
Abstract: The underwater sensor network is an important measure to realize the intelligence of marine environment monitoring, the propagation rate of the hydroacoustic signal is nonlinear and unstable due to the physical characteristics of seawater, which makes the ranging method based on hydroacoustic signal inevitably have a large-ranging error. This paper proposes an underwater localization algorithm based on convolutional long and short-term memory neural network (ConvLSTM) corrected ranging, which uses the neural network constructed based on ConvLSTM to extract features from the historical data of the marine environment and make a hierarchical prediction of the hydroacoustic signal speed rate, to reduce the ranging error. The corrected ranging values are substituted into the improved LSM-based localization algorithm, which can further improve the localization accuracy of unknown nodes underwater. The simulation experimental results show that the algorithm can significantly reduce the localization error and has high localization accuracy and feasibility.
Key words: underwater positioning algorithm; sound speed prediction; ConvLSTM; underwater sensor network
收稿日期: 2025-03-10
基金项目: 国家自然科学基金项目(52571384).
作者简介: 张 瑞(1998—),男,硕士研究生. * 通信联系人.