一种粗精匹配结合的重力辅助惯性导航算法  

A gravity aided inertial navigation algorithm based on the rough/fine matching combination

作　　者：张锦柏 肖云 陈垲宁[1,2] 曹杰 洪晓东 王宇康 ZHANG JinBai;XIAO Yun;CHEN KaiNing;CAO Jie;HONG XiaoDong;WANG YuKang(School of Geological Engineering and Geomatics,Chang'an University,Xi’an 710054,China;National Key Laboratory of Spatiotemporal Datum,Xi’an 710054,China;Xi’an Research Institute of Surveying and Mapping,Xi’an 710054,China)

机构地区：[1]长安大学地质工程与测绘学院,西安710054 [2]空间基准全国重点实验室,西安710054 [3]西安测绘研究所,西安710054

出　　处：《地球物理学报》2025年第3期788-801,共14页Chinese Journal of Geophysics

基　　金：国家自然科学基金(41374083)资助。

摘　　要：重力匹配导航是目前水下纠正惯性导航累积漂移误差的有效方式,既符合水下导航隐蔽性要求,又可保证潜器高精度导航定位.针对相关极值算法实时性及精度较差的问题,本文提出一种粗精匹配结合的重力辅助惯性导航算法.基于惯导误差随时间累积与其在短时内漂移误差不大的特点,约束并控制匹配点数量与质量.对观测点集合进行粗匹配,筛选合适结果作为精匹配的修正基础;结合最近点迭代算法对粗匹配轨迹进行旋转和尺度修正,补偿惯导误差,得到精匹配轨迹,并与传统匹配算法对比分析.仿真结果表明:该算法粗精结合的优势能稳定提升匹配精度,可排除大量误匹配点,减少搜索时间近40%,经迭代变换后精度由千米级提升到百米级,显著提高了算法精度与实时性.Gravity matching navigation is an effective method for correcting the cumulative drift error of inertial navigation underwater,meeting the requirements of stealth and ensuring high-precision navigation and positioning for submersibles.To address the issues of poor real-time performance and accuracy in correlation extremum algorithms,this paper proposes a gravity-aided inertial navigation algorithm based on the rough/fine matching combination.Leveraging the characteristic that inertial navigation errors accumulate over time but exhibit minimal drift in the short term,the algorithm constrains and controls the quantity and quality of matching points.Rough matching is performed on the observation point set,and suitable results are selected as the basis for fine matching correction.The iterative closest point algorithm is then used to adjust the rotation and scale of the rough-matched track,compensating for inertial navigation errors and yielding a fine-matched track.Comparative analysis with traditional matching algorithms demonstrates that the proposed method significantly enhances matching accuracy by effectively excluding mismatched points and reducing search time by nearly 40%.After iterative transformation,the accuracy improves from the kilometer level to the hundred-meter level,markedly enhancing both the precision and real-time performance of the algorithm.

关 键 词：粗精结合 重力辅助导航 惯性导航 均方差匹配 迭代变换 

分 类 号：P223[天文地球—大地测量学与测量工程]