基于组合相似度的实时矢量地磁轮廓匹配算法

Real-time vector geomagnetic contour matching algorithm based on combined similarity

关键词:  实时矢量地磁轮廓匹配算法  组合相似度度量  多尺度搜索策略  地磁导航  自适应权重机制  实时性能

Key words:  Real-time vector geomagnetic contour matching algorithm  Combined similarity metics  Multi-scale search strategy  Geomagnetic navigation  Adaptive weighting mechanism  Real-time performance

摘要:

为了提高地磁导航系统的定位精度和实时性能，提出了一种基于组合相似度的实时矢量地磁轮廓匹配算法。该算法引入了矢量地磁特征和组合相似度度量，构建了多维度相似度模型，包括方向相似度和强度相似度。方向相似度采用余弦相似度计算，强度相似度则使用归一化欧氏距离计算。算法采用了高效的多层金字塔搜索策略，实现了实时矢量地磁匹配，在保证匹配精度的同时显著提高了计算效率。通过仿真实验对算法进行了全面评估。结果表明，在最优参数组合下，算法的均方根误差（RMSE）为52.48 m，更新频率达到2.87 Hz，明显优于传统地磁轮廓匹配（MAGCOM）算法。在噪声强度从0 nT增加到5 nT时，该算法的RMSE从52.48 m增加到399.56 m，而MAGCOM算法则从444.89 m增加到890.66 m，展现出更强的抗干扰能力。通过调整参数，算法的更新频率可进一步提高到4.06 Hz。研究结果表明，该算法在保证高精度定位的同时，满足实时处理要求，为地磁辅助导航系统提供了高效可靠的解决方案。

Abstract:

A real-time vector geomagnetic contour matching algorithm based on combined similarity is proposed to improve the positioning accuracy and real-time performance of a geomagnetic navigation system. In this algorithm, the vector geomagnetic features and combined similarity metrics are introduced, and a multi-dimensional similarity model, including direction and intensity similarities, is constructed. Directional similarity is quantified by using cosine similarity, while intensity similarity is calculated by using normalized Euclidean distance. The algorithm uses an efficient multi-layer pyramid search strategy to achieve real-time vector geomagnetic matching, significantly improving computational efficiency while ensuring matching accuracy. A comprehensive evaluation of the algorithm is carried out through simulation experiments. The results show that under the optimal parameter combination, the algorithm achieves a root mean square error (RMSE) of 52.48 m and an update frequency of 2.87 Hz, significantly outperforming the traditional magnetic contour matching (MAGCOM) algorithm. As the noise intensity increases from 0 nT to 5 nT, the RMSE of this algorithm increases from 52.48 m to 399.56 m, while that of MAGCOM increases from 444.89 m to 890.66 m, demonstrating stronger anti-interference capability. By adjusting the parameters, the update frequency of the algorithm can be further increased to 4.06 Hz. The research results show that this algorithm meets real-time processing requirements while ensuring high-precision positioning, providing an efficient and reliable solution for geomagnetic-assisted navigation systems.