基于三维格网误差建模的GNSS/IMU抗差自适应定位算法

盛琪; 孙蕊; 何雨霖; 张珩瑜

doi:10.13700/j.bh.1001-5965.2024.0169

基于三维格网误差建模的GNSS/IMU抗差自适应定位算法

doi: 10.13700/j.bh.1001-5965.2024.0169

南京航空航天大学民航学院，南京 211106

基金项目:

江苏省自然科学基金(BK20211569)

详细信息

通讯作者:
E-mail：rui.sun@nuaa.edu.cn

中图分类号: V324；P228；V249.3
计量
- 文章访问数: 380
- HTML全文浏览量: 99
- PDF下载量: 11
- 被引次数: 0
出版历程
- 收稿日期: 2024-03-26
- 录用日期: 2024-05-31
- 网络出版日期: 2024-08-30
- 整期出版日期: 2026-05-26

A robust adaptive positioning algorithm for GNSS/IMU based on 3D grid error modeling

College of Civil Aviation，Nanjing University of Aeronautics and Astronautics，Nanjing 211106，China

Funds:

Natural Science Foundation of Jiangsu Province (BK20211569)

More Information

Corresponding author: E-mail：rui.sun@nuaa.edu.cn

摘要

摘要:
城市复杂环境中高大建筑会造成全球卫星导航系统（GNSS）、信号非视距传播（NLOS）和多径干扰（MI），影响智能交通定位精度；现有二维格网多径建模方法存在高程精度不足、测量噪声协方差调参策略简单的缺陷。提出一种基于三维格网误差建模的GNSS/惯性测量单元（IMU）抗差自适应定位算法：在现有二维格网的基础上，划分高度空间，进一步实现精细化建模，在多径误差预测阶段，通过格网中心匹配方法缓解错误匹配导致的模型预测误差；基于多径误差预测值，提出滤波模型选择策略，并结合抗差理论提出抗差阈值动态调节策略，达到对测量噪声协方差的环境自适应更新，能有效提升城市复杂场景GNSS/IMU组合导航定位精度。车载实验结果表明：所提算法相较传统GNSS/IMU紧组合算法和传统GNSS/IMU抗差自适应算法，定位精度分别提升了48.43%和31.48%，相较二维格网辅助的GNSS/IMU抗差自适应算法提升了27.57%。
- 多径误差建模 /
- 抗差滤波 /
- 自适应滤波 /
- GNSS/IMU组合导航 /
- 非视距传播
Abstract:
In complex urban environments, tall buildings cause Global Navigation Satellite System (GNSS) signals to suffer from Non-Line-of-Sight (NLOS) propagation and Multipath Interference (MI), which degrade the positioning accuracy of intelligent transportation systems. The existing two-dimensional grid-based multipath modeling method has shortcomings including insufficient precision in the height direction and an overly simplistic adjustment strategy for measurement noise covariance. This article proposes a GNSS/inertial measurement unit (IMU) robust adaptive filter algorithm based on 3D grid error modeling. By dividing the height space on the basis of the existing 2D grid, fine modeling can be further achieved. In the stage of multipath error prediction, the grid-center-matching method is used to alleviate the model prediction error caused by incorrect matching. Then, we propose a filtering model selection strategy based on the multipath error predictions. Moreover, according to the robust theory, we propose a robust threshold dynamic adjustment strategy to update the measurement noise covariance adaptively. The positioning performance of GNSS/IMU integrated navigation in complex urban environments can be significantly improved by the proposed algorithm. The 3D positioning accuracy of the suggested algorithm has improved by 27.57% when compared to the 2D grid assisted GNSS/IMU robust adaptive algorithm and by 48.44% and 31.51%, respectively, when compared to the traditional GNSS/IMU tight combination algorithm and the traditional GNSS/IMU robust adaptive algorithm, according to the results of urban environment vehicle experiments.
- pseudorange error modeling /
- robust filtering /
- adaptive filtering /
- GNSS/IMU integrated navigation /
- non-line-of-sight

HTML全文

图 1 算法框架

Figure 1. Algorithm framework

下载: 全尺寸图片幻灯片

图 2 城市区域二维格网划分示意图

Figure 2. 2D grid division of urban areas diagram

下载: 全尺寸图片幻灯片

图 3 高度划分示意图

Figure 3. Diagram of the division in height

下载: 全尺寸图片幻灯片

图 4 袋装回归树算法流程

Figure 4. Flowchart of the ensemble bagged regression tree algorithm

下载: 全尺寸图片幻灯片

图 5 三维格网辅助的抗差自适应滤波算法框图

Figure 5. Flowchart of the 3D grid-aided robust adaptive filtering algorithm

下载: 全尺寸图片幻灯片

图 6 二维格网匹配示意图

Figure 6. Diagram of 2D grid matching

下载: 全尺寸图片幻灯片

图 7 动态调节 $ {c}_{0} $示意图

Figure 7. Diagram of dynamic adjustment $ {c}_{0} $

下载: 全尺寸图片幻灯片

图 8 实验环境

Figure 8. Experiment environment

下载: 全尺寸图片幻灯片

图 9 车载实验平台

Figure 9. Vehicle test platform

下载: 全尺寸图片幻灯片

图 10 卫星数

Figure 10. Number of satellites

下载: 全尺寸图片幻灯片

图 11 PDOP值

Figure 11. PDOP values

下载: 全尺寸图片幻灯片

图 12 4种算法定位轨迹对比

Figure 12. Comparison of positioning trajectories of the four algorithms

下载: 全尺寸图片幻灯片

图 13 北东地定位误差

Figure 13. Positioning error in NED

下载: 全尺寸图片幻灯片

图 14 水平和三维定位误差

Figure 14. Horizontal and 3D positioning error

下载: 全尺寸图片幻灯片

表 1 模型训练样本数量

Table 1. Number of training samples of the models

星座	平均样本数量	最大样本数量	最少样本数量
GPS	979	20 563	200
BDS	1429	31 901	200

下载: 导出CSV

表 2 算法描述

Table 2. Algorithm description

算法	算法描述
算法1	1）传统卡尔曼滤波 2）GNSS/IMU紧组合
算法2	1）传统抗差自适应滤波 2）GNSS/IMU紧组合
算法3	1）二维格网误差建模辅助抗差自适应滤波选择 2）二维格网误差建模辅助权函数阈值动态调节 3）GNSS/IMU紧组合
本文算法	1）三维格网误差建模辅助抗差自适应滤波选择 2）三维格网误差建模辅助权函数阈值动态调节 3）GNSS/IMU紧组合

下载: 导出CSV

表 3 4种算法定位精度对比

Table 3. Positioning accuracy comparison of the four algorithms

方向	均方根误差/m				提升比例/%
方向	算法1	算法2	算法3	本文算法	算法2较算法1	算法3较算法1	本文算法较算法1	本文算法较算法2	本文算法较算法3
北	9.35	6.97	7.76	2.62	25.45	17	71.98	62.41	66.24
东	12.67	8.05	6.24	2.87	36.464	50.75	77.35	64.35	54.01
地	10.91	9.74	9.31	9.08	10.72	14.7	16.77	6.78	2.47
水平	15.75	10.64	9.96	3.88	32.40	36.76	75.37	63.53	61.04
3D	19.16	14.42	13.64	9.88	24.74	28.81	48.43	31.48	27.57

下载: 导出CSV

参考文献(24)

[1]	MACGOUGAN G, LACHAPELLE G, KLUKAS R, et al. Performance analysis of a stand-alone high-sensitivity receiver[J]. GPS Solutions, 2002, 6(3): 179-195.
[2]	赵泽方, 杜平, 邵玉. 小型化3D扼流圈天线设计[J]. 中国科技论文, 2016, 11(14): 1605-1608. ZHAO Z F, DU P, SHAO Y. Design of a miniaturized 3D choke ring antenna[J]. China Sciencepaper, 2016, 11(14): 1605-1608(in Chinese).
[3]	陆思丞, 王昕, 张胜辉. 一种小型化低剖面天线扼流圈的设计[C]//2019年全国微波毫米波会议论文集. 2019: 191-194. LU S C, WANG X, ZHANG S H. Design of a miniaturized low-profile antenna choke[C]//Proceedings of the 2019 National Conference on Microwave and Millimeter Waves. Guangzhou: Chinese Institute of Electronics, 2019: 191-194.
[4]	BRENNEMAN M, MORTON J, YANG C, et al. Mitigation of GPS multipath using polarization and spatial diversities[C]//Proceedings of the 20th International Technical Meeting of the Satellite Division of The Institute of Navigation. 2007: 1221-1229.
[5]	JIANG Z Y, GROVES P D. NLOS GPS signal detection using a dual-polarisation antenna[J]. GPS solutions, 2014, 18(1): 15-26.
[6]	TRANQUILLA J M, CARR J P, AL-RIZZO H M. Analysis of a choke ring groundplane for multipath control in Global Positioning System (GPS) applications[J]. IEEE Transactions on Antennas and Propagation, 1994, 42(7): 905-911.
[7]	王尔申, 张淑芳, 张芝贤. GPS接收机抗多径技术研究现状与趋势[J]. 电讯技术, 2011, 51(1): 114-119. WANG E S, ZHANG S F, ZHANG Z X. Research status and trend of mitigating multipath effects technology for GPS receiver[J]. Telecommunication Engineering, 2011, 51(1): 114-119(in Chinese).
[8]	SUN R, WANG G, ZHANG W, et al. A gradient boosting decision tree based GPS signal reception classification algorithm[J] Applied Soft Computing, 2020, 86: 105942.
[9]	QIN F, FU L X, WANG Y Q, et al. A bagging tree-based pseudorange correction algorithm for global navigation satellite system positioning in foliage canyons[J]. International Journal of Distributed Sensor Networks, 2021, 17(5): 155014772110167.
[10]	SUN R, FU L, CHENG Q, et al. Resilient pseudorange error prediction and correction for GNSS positioning in urban areas[J]. IEEE Internet of Things Journal, 2023, 10(11): 9979-9988.
[11]	ZHANG Z X, SUN R. Improving loosely coupled GNSS/IMU fusion performance with pseudorange error prediction in urban areas[M]//Advances in Guidance, Navigation and Control. Singapore: Springer Nature, 2023: 2850-2860.
[12]	SUN R, ZHANG Z X, CHENG Q, et al. Pseudorange error prediction for adaptive tightly coupled GNSS/IMU navigation in urban areas[J]. GPS Solutions, 2021, 26(1): 28.
[13]	王坚, 刘超, 高井祥, 等. 基于抗差EKF的GNSS/INS紧组合算法研究[J]. 武汉大学学报(信息科学版), 2011, 36(5): 596-600. WANG J, LIU C, GAO J X, et al. GNSS/INS tightly coupled navigation model based on robust EKF[J]. Geomatics and Information Science of Wuhan University, 2011, 36(5): 596-600(in Chinese).
[14]	吴富梅, 杨元喜. 基于小波阈值消噪自适应滤波的GPS/INS组合导航[J]. 测绘学报, 2007, 36(2): 124-128. WU F M, YANG Y X. GPS/INS integrated navigation by adaptive filtering based on wavelet threshold de-noising[J]. Acta Geodaetica et Cartographica Sinica, 2007, 36(2): 124-128(in Chinese).
[15]	YANG Y, HE H, XU G. Adaptively robust filtering for kinematic geodetic positioning[J]. Journal of Geodesy, 2001, 75(2): 109-116.
[16]	杨元喜, 任夏, 许艳. 自适应抗差滤波理论及应用的主要进展[J]. 导航定位学报, 2013, 1(1): 9-15. YANG Y X, REN X, XU Y. Main progress of adaptively robust filter with applications in navigation[J]. Journal of Navigation and Positioning, 2013, 1(1): 9-15(in Chinese).
[17]	NING Y, WANG J, HAN H, et al. An optimal radial basis function neural network enhanced adaptive robust Kalman filter for GNSS/INS integrated systems in complex urban areas[J]. Sensors, 2018, 18(9): 3091.
[18]	GAO B B, HU G G, ZHONG Y M, et al. Cubature Kalman filter with both adaptability and robustness for tightly-coupled GNSS/INS integration[J]. IEEE Sensors Journal, 2021, 21(13): 14997-15011.
[19]	SONG B, MA X D, ZHANG S B. Analysis of BDS Satellite Orbit Periodicity[J]. GNSS World of China, 2016, 41(5): 47-50.
[20]	YOZEVITCH R, BEN MOSHE B, WEISSMAN A. A robust GNSS LOS/NLOS signal classifier[J]. Navigation, 2016, 63(4): 429-442.
[21]	SAEED M S, MUSTAFA M W, SHEIKH U U, et al. Ensemble bagged tree based classification for reducing non-technical losses in multan electric power company of Pakistan[J]. Electronics, 2019, 8(8): 860.
[22]	MOUSAVI R, EFTEKHARI M. A new ensemble learning methodology based on hybridization of classifier ensemble selection approaches[J]. Applied Soft Computing, 2015, 37: 652-666.
[23]	ATIA M M, WASLANDER S L. Map-aided adaptive GNSS/IMU sensor fusion scheme for robust urban navigation[J]. Measurement, 2019, 131: 615-627.
[24]	WANG L, GROVES P D, ZIEBART M. Multi-constellation GNSS performance evaluation for urban canyons using large virtual reality city models[J]. Journal of Navigation, 2012, 65(3): 459-476.