无人机气压计测高模型精化及GNSS/SINS组合定位增强

doi:10.11947/j.AGCS.2025.20250276

摘要/Abstract

摘要：

无人机导航除了GNSS/SINS模块，通常还搭载了低成本微电子机械系统（MEMS）气压计，现有的气压计测高与GNSS基准不一致，且易受气流、风力等因素影响，随机误差较大，难以可靠地应用在无人机导航中。基于此，本文提出了一种基于气压计测高误差精细建模的无人机GNSS/SINS组合导航增强方法。具体地，在经典GNSS/SINS的15维状态基础上，引入气压计偏置状态进行随机游走参数化建模，鉴于气压计误差波动大，针对旋翼无人机气压计测高随机误差进行显著相关性分析，提出了一种基于运动状态的气压计测高误差预测的神经网络模型，建立惯性测量单元（IMU）输出与气压计随机误差之间的映射关系，作为气压计增强定位的随机模型优化的基础。通过山区无人机飞行试验，结果表明，相较于传统方法，气压计测高误差精细化建模并与GNSS/SINS进行扩展卡尔曼滤波（EKF）融合在开阔大机动场景下，天向定位精度提升了15.20%；在峡谷遮挡环境中，天向定位精度提升了37.74%；针对10 s GNSS信号拒止的情况，天向定位精度则提升44.20%，证明在山区等复杂场景下本文方法具有较好的稳健性及更高的定位精度。

关键词: GNSS/SINS, 无人机, 气压计, 卷积神经网络, 高度约束

Abstract:

In addition to the GNSS/SINS module, UAV navigation systems are often equipped with low-cost micro-electro-mechanical systems (MEMS) barometers. However, existing barometric height measurements suffer from inconsistency with GNSS references and are susceptible to disturbances such as airflow and wind, resulting in significant random errors that limit their reliable application in UAV navigation. To address this issue, this paper proposes a novel enhanced method for integrated GNSS/SINS navigation of UAVs based on refined error modeling of barometric height measurements. Specifically, on the basis of the classical 15-state GNSS/SINS model, a barometer bias state is introduced and parameterized using a random walk model. Furthermore, considering the high volatility of barometric errors, a significant correlation analysis is conducted on the random errors of barometric height measurements in rotary-wing UAVs. A neural network model based on motion state is proposed to predict barometric height errors, establishing a mapping relationship between inertial measurement unit (IMU) outputs and barometric random errors, which serves as the foundation for optimizing the stochastic model in barometer-enhanced positioning. Experimental results from UAV flight tests in mountainous areas demonstrate that the proposed refined barometric error modeling method, when integrated with GNSS/SINS using extended Kalman filter (EKF), significantly improves vertical positioning accuracy compared to conventional methods. In open areas with high-dynamic maneuvers, the vertical positioning accuracy is improved by 15.20%; in canyon environments with signal occlusion, it is improved by 37.74%; and under 10-second GNSS-denied conditions, the accuracy is improved by 44.20%. These results confirm that the proposed method offers strong robustness and higher positioning accuracy in complex scenarios such as mountainous regions.

Key words: GNSS/SINS, UAV, barometer, convolutional neural network, altitude constraint

中图分类号:

P228

宋瀚昀, 李昕, 黄观文, 李航. 无人机气压计测高模型精化及GNSS/SINS组合定位增强[J]. 测绘学报, 2025, 54(11): 1980-1991.

Hanyun SONG, Xin LI, Guanwen HUANG, Hang LI. Refinement of UAV barometer altimetry model and GNSS/SINS integrated positioning enhancement[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(11): 1980-1991.

图/表 12

图1

图2

图3

图4

表1

图5

图6

图7

图8

图9

图10

表2

参考文献 32

[1]	ZHANG Guohao, HSU L T. Intelligent GNSS/INS integrated navigation system for a commercial UAV flight control system[J]. Aerospace Science and Technology, 2018, 80: 368-380.
[2]	GUERRA A, DARDARI D, DJURIC P M. Joint indoor localization and navigation of UAVs for network formation control[C]//Proceedings of 2018 Asilomar Conference on Signals, Systems, and Computers. Pacific Grove: IEEE, 2018: 13-19.
[3]	CHIANG Kaiwei, CHANG H W, LI Yuhua, et al. Assessment for INS/GNSS/odometer/barometer integration in loosely-coupled and tightly-coupled scheme in a GNSS-degraded environment[J]. IEEE Sensors Journal, 2020, 20(6): 3057-3069.
[4]	ZHU Ni, MARAIS J, BETAILLE D, et al. GNSS position integrity in urban environments: a review of literature[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(9): 2762-2778.
[5]	SCHEIERMANN S. MEMS传感器——无人机的核心[J]. 传感器世界, 2019, 25(5): 36-40.
	SCHEIERMANN S. MEMS sensors—hearts of UAVs[J]. Sensor World, 2019, 25(5): 36-40.
[6]	SIRIKONDA S, PARAYITAM L, HABLANI H B. Performance enhancement of NavIC with low-cost MEMS-IMU, barometer, and magnetometer[J]. IEEE Access, 2025(13): 17303-17324.
[7]	SABATINI A M, GENOVESE V. A sensor fusion method for tracking vertical velocity and height based on inertial and barometric altimeter measurements[J]. Sensors, 2014, 14(8): 13324-13347.
[8]	YEN S W, VAN GRAAS F, DE HAAG M U. Positioning with two satellites and known receiver clock, barometric pressure and radar elevation[J]. GPS Solutions, 2016, 20(4): 885-899.
[9]	CHEN H, WU W, ZHANG S, et al. A GNSS/LiDAR/IMU pose estimation system based on collaborative fusion of factor map and filtering[J]. Remote Sensing, 2023, 15(3): 790.
[10]	李彦杰, 杨元喜, 何海波. 附加约束条件对GNSS/INS组合导航结果的影响分析[J]. 武汉大学学报(信息科学版), 2017, 42(9): 1249-1255.
	LI Yanjie, YANG Yuanxi, HE Haibo. Effects analysis of constraints on GNSS/INS integrated navigation[J]. Geomatics & Information Science of Wuhan University, 2017, 42(9): 1249-1255.
[11]	SON Y, OH S. A barometer-IMU fusion method for vertical velocity and height estimation[C]//Proceedings of 2015 IEEE SENSORS. Busan: IEEE, 2015: 1-4.
[12]	BAO Le, LI Kai, LEE J, et al. An enhanced indoor three-dimensional localization system with sensor fusion based on ultra-wideband ranging and dual barometer altimetry[J]. Sensors, 2024, 24(11): 3341.
[13]	SHIN E H. Estimation techniques for low-cost inertial navigation[D]. Calgary: University of Calgary, 2005.
[14]	TANIGAWA M, LUINGE H, SCHIPPER L, et al. Drift-free dynamic height sensor using MEMS IMU aided by MEMS pressure sensor[C]//Proceedings of the 5th Workshop on Positioning, Navigation and Communication. Hannover: IEEE, 2008: 191-196.
[15]	LI Zehua, SHANG Junna, YAO Ziyang. Indoor integrated navigation on PDR/Wi-Fi/barometer via factor graph with local attention[J]. IEEE Sensors Journal, 2023, 23(22): 27613-27623.
[16]	PARVIAINEN J, KANTOLA J, COLLIN J. Differential barometry in personal navigation[C]//Proceedings of 2008 IEEE/ION Position, Location and Navigation Symposium. Monterey: IEEE, 2008: 148-152.
[17]	黄鹤, 刘一恒, 赵熙, 等. 多层多源信息融合旋翼无人机测高算法[J]. 中国惯性技术学报, 2018, 26(3): 316-322, 329.
	HUANG He, LIU Yiheng, ZHAO Xi, et al. Multi-layer multi-source information fusion algorithm for hovering UAV height measurement[J]. Journal of Chinese Inertial Technology, 2018, 26(3): 316-322, 329.
[18]	ZHANG Jieying, EDWAN E, ZHOU Junchuan, et al. Performance investigation of barometer aided GPS/MEMS-IMU integration[C]//Proceedings of 2012 IEEE/ION Position, Location and Navigation Symposium. Myrtle Beach: IEEE, 2012: 598-604.
[19]	PARK J, LEE Dongsun, PARK C. Implementation of vehicle navigation system using GNSS, INS, odometer and barometer[J]. Journal of Positioning, Navigation, and Timing, 2015, 4(3): 141-150.
[20]	XU Dacheng, YANG Zhimei, ZHAO Heming, et al. A temperature compensation method for MEMS accelerometer based on LM_BP neural network[C]//Proceedings of 2016 IEEE SENSORS. Orlando: IEEE, 2016: 1-3.
[21]	LIU Yaohua, LUO Qingsong, ZHOU Yimin. Deep learning-enabled fusion to bridge GPS outages for INS/GPS integrated navigation[J]. IEEE Sensors Journal, 2022, 22(9): 8974-8985.
[22]	ZHANG Shuning, WANG Ershen, WANG Yongkang, et al. A low-cost UAV swarm relative positioning architecture based on BDS/barometer/UWB[J]. IEEE Sensors Journal, 2024, 24(23): 39659-39668.
[23]	马旭, 程咏梅, 郝帅. 面向无人机高度融合估计的自适应S滤波方法[J]. 中国惯性技术学报, 2013, 21(5): 604-608.
	MA Xu, CHENG Yongmei, HAO Shuai. Adaptive S filtering method for UAV height fusion estimation[J]. Journal of Chinese Inertial Technology, 2013, 21(5): 604-608.
[24]	ZHAO Xiaoxi, MENG Yue, QI Fei, et al. A vertical channel-enhanced fusion method based on RINS and barometric altimeter for UAVs in GNSS denial environments[J]. IEEE Transactions on Instrumentation and Measurement, 2023, 72: 1-12.
[25]	CHE Yifan, DONG Jiuxiang. A real-time, robust visual-inertial navigation system tightly coupled with GNSS and barometer[J]. IEEE Sensors Letters, 2024, 8(6): 1-4.
[26]	CAO Yin, XU Wenyuan, LIN Bo, et al. A method for temperature error compensation in fiber-optic gyroscope based on machine learning[J]. Optik, 2022, 256: 168765.
[27]	CHEN Hua, AGGARWAL P, TAHA T M, et al. Improving inertial sensor by reducing errors using deep learning methodology[C]//Proceedings of 2018 IEEE National Aerospace and Electronics Conference. Dayton: IEEE, 2018: 197-202.
[28]	JIANG Changhui, CHEN Shuai, CHEN Yuwei, et al. A MEMS IMU de-noising method using long short term memory recurrent neural networks (LSTM-RNN)[J]. Sensors, 2018, 18(10): 3470.
[29]	LI Xin, LI Hanxu, HUANG Guanwen, et al. Non-holonomic constraint (NHC)-assisted GNSS/SINS positioning using a vehicle motion state classification (VMSC)-based convolution neural network[J]. GPS Solutions, 2023, 27(3): 144.
[30]	李晗旭, 李昕, 黄观文, 等. 速度分类的LSTM神经网络辅助GNSS/SINS车载定位方法[J]. 武汉大学学报(信息科学版), 2025, 50(7): 1311-1320.
	LI Hanxu, LI Xin, HUANG Guanwen, et al. LSTM neural network assisted GNSS/SINS vehicle positioning based on speed classification[J]. Geomatics and Information Science of Wuhan University, 2025, 50(7): 1311-1320.
[31]	WANG Jian, HAN Houzeng, LIU Fei, et al. Performance analysis of GNSS/MIMU tight fusion positioning model with complex scene feature constraints[J]. Journal of Geodesy and Geoinformation Science, 2021, 4(2): 1-13.
[32]	李昕, 孟硕林, 黄观文, 等. GNSS/SINS定位稳健SE2(3)-EKF方法[J]. 测绘学报, 2023, 52(10): 1640-1649. DOI: 10.11947/j.AGCS.2023.20220526.
	LI Xin, MENG Shuolin, HUANG Guanwen, et al. Robust GNSS/SlNS positioning based on the SE2(3)-EKF framework[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(10): 1640-1649. DOI: 10.11947/j.AGCS.2023.20220526.

参数	数值
陀螺零偏不稳定性/（（°）/h）	10
陀螺随机游走/（（°）/）	0.7
加速度计稳定性/mg	0.1
加速度计随机游走/（mg/）	0.5
气压计压力分辨率/mbar	0.02
气压计测高标称精度/m	0.2

场景	轴向	RMSE/m			U向提升/（%）
场景	轴向	GNSS/SINS	GNSS/SINS/BARO（固定R_h）	GNSS/SINS/BARO（BHE-Net）	U向提升/（%）
开阔机动	E	0.547 7	0.510 9	0.464 4	15.20
	N	0.595 3	0.590 9	0.507 5
	U	0.523 9	0.459 1	0.410 9
	3D	0.963 7	0.906 1	0.801 3
峡谷遮挡	E	0.754 0	0.702 6	0.678 4	37.74
	N	1.287 8	1.150 0	0.838 1
	U	0.283 8	0.182 3	0.176 7
	3D	1.519 1	1.360 0	1.092 6
GNSS拒止（10 s）	E	1.474 6	0.845 4	0.780 9	44.20
	N	3.524 1	1.766 2	1.417 0
	U	0.674 2	0.379 3	0.376 8
	3D	3.879 8	1.994 5	1.661 1