| 基于船载5G增强型系统架构的目标轨迹融合 |
| Target Trajectory Fusion Based on an Enhanced Shipborne 5G System Architecture |
| 投稿时间：2025-07-16 修订日期：2025-12-16 |
| DOI： |
| 中文关键词: 5G通信 AIS APRA 信息融合 边缘计算 低轨卫星组网 |
| 英文关键词:5G communication AIS ARPA information fusion edge computing LEO Satellite Networking |
| 基金项目:基于船载5G边缘计算架构的海域监控综合应用平台研究（ZDYF2024GXJS034） |
|
| 摘要点击次数: 111 |
| 全文下载次数: 0 |
| 中文摘要: |
| 本文提出一种基于船载5G增强型系统架构下船舶自动识别系统（Automatic Identification System，AIS）与雷达自动标绘仪（Automatic Radar Plotting Aid，ARPA）目标轨迹动态信息融合算法，旨在解决船舶航行中AIS与ARPA目标轨迹信息融合的实时性与精确度问题。通过构建半实物仿真平台，不仅集成了5G网络模拟与边缘计算架构，还设计了自适应传输策略。改进的联合概率数据关联（Joint Probabilistic Data Association，JPDA）与运动预测模型相结合，使得航向角误差在目标交叉场景下成功降至1.87°，在恶劣海况场景下，最优改进率达50.32%。这些结果通过多次测试得到了证实，显示出系统在复杂场景下的优越性能。通过将5G通信技术与边缘计算相结合，在典型港口场景下将端到端时延压缩至129.70ms，完全满足国际海事组织（International Maritime Organization，IMO）避碰系统的实时性要求。所提出的动态权重优化机制表现出色，使航向估计方差降低45%以上，即使在密集目标场景下仍能保持2.67m的定位精度，相较传统方法提升47.85%。这些数据不仅在实验中得到了支持，而且在实际应用中具有重要的参考价值。 |
| 英文摘要: |
| Aiming to address the real-time performance and accuracy challenges in fusing AIS and ARPA target trajectory data during ship navigation, a dynamic information fusion algorithm for AIS (Automatic Identification System) and ARPA(Automatic Radar Plotting Aid) target trajectories under an enhanced ship borne 5G system architectures proposed. By constructing a hardware-in-the-loop simulation platform, it not only integrates 5G network simulation and edge computing architecture but also designs an adaptive transmission strategy. The improved JPDA(Joint Probabilistic Data Association) combined with a motion prediction model successfully reduces heading angle errors to 1.87° in target-crossing scenarios and achieves an optimal improvement rate of 50.32% under harsh sea conditions. these results, validated through multiple tests, demonstrate the system's superior performance in complex scenarios. By integrating 5G communication technology with edge computing, the end-to-end latency is compressed to 129.70ms in typical port scenarios, fully meeting the real-time requirements of IMO(International Maritime Organization) collision avoidance systems. The proposed dynamic weight optimization mechanism performs excellently, reducing the heading estimation variance by over 45% and maintaining a positioning accuracy of 2.67m even in dense target scenarios, which is a 47.85% improvement compared to traditional methods. These findings are not only experimentally validated but also hold significant reference value for practical applications. |
| 查看/发表评论 下载PDF阅读器 |
