基于速度分类的LSTM神经网络辅助GNSS/SINS车载定位方法

李晗旭, 李昕, 黄观文, 张勤, 陈世鹏

李晗旭, 李昕, 黄观文, 张勤, 陈世鹏. 基于速度分类的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). DOI: 10.13203/j.whugis20230061
引用本文: 李晗旭, 李昕, 黄观文, 张勤, 陈世鹏. 基于速度分类的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). DOI: 10.13203/j.whugis20230061
LI Hanxu, LI Xin, HUANG Guanwen, ZHANG Qin, CHEN Shipeng. LSTM Neural Network Assisted GNSS/SINS Vehicle Positioning Based on Speed Classification[J]. Geomatics and Information Science of Wuhan University. DOI: 10.13203/j.whugis20230061
Citation: LI Hanxu, LI Xin, HUANG Guanwen, ZHANG Qin, CHEN Shipeng. LSTM Neural Network Assisted GNSS/SINS Vehicle Positioning Based on Speed Classification[J]. Geomatics and Information Science of Wuhan University. DOI: 10.13203/j.whugis20230061

基于速度分类的LSTM神经网络辅助GNSS/SINS车载定位方法

基金项目: 

国家自然科学基金委重大科研仪器项目(42127802);国家自然科学基金(42004023);科技部重点研发计划(2021YFB2600603)。

详细信息
    作者简介:

    李晗旭,硕士,主要从事GNSS/SINS组合导航。957188494@qq.com

    通讯作者:

    李昕,博士,副教授。lixin2017@chd.edu.cn

  • 中图分类号: P228

LSTM Neural Network Assisted GNSS/SINS Vehicle Positioning Based on Speed Classification

  • 摘要: 在车载GNSS/SINS组合导航中,为了增强定位效果,车辆非完整性约束(NHC)是常用的增强方法。通过应用机器学习的方法来建立IMU输出与NHC伪观测量之间的复杂映射关系,并直接在观测域对NHC伪观测量的大小进行调整,可以提升传统NHC方法的约束效果。现有的机器学习方法没有考虑车辆运动状态影响,导致NHC预测精度和可靠性不高。最新研究表明机器学习可以预测车辆的前向速度,即虚拟里程计(ODO)。然而,当前研究主要是将预测虚拟NHC和虚拟ODO分开讨论,没有充分挖掘二者之间的耦合关系以及三维速度对车辆进行全约束的能力。因此,本文的研究针对这一问题进行了深入探讨,提出了一种基于车辆速度分类的LSTM神经网络,用于预测车辆的三维速度,并采用三维速度约束新息来自适应调整其方差域。为了验证本文方法的有效性,进行了车载GNSS/SINS组合导航的实验测试。根据实验结果显示,本文方法在前向速度预测方面的平均精度约为0.4 m/s,在侧向和垂向速度预测方面平均精度约为2 cm/s,此外,在模拟GNSS信号失锁460秒的情况下,相较于惯导推算结果,本文方法在三维速度约束下的水平定位精度改善了99.40%。
    Abstract: Vehicular Non-holonomic constraint (NHC) is a commonly used enhancement technique in combined vehicle-mounted GNSS/SINS navigation in order to enhance the positioning effect. The constraint effect of traditional NHC methods is enhanced by applying machine learning methods to establish a complex mapping relationship between IMU outputs and NHC pseudo-observations, and to adjust the size of NHC pseudo-observations directly in the observation domain. Existing machine learning methods do not consider the influence of vehicle motion state, resulting in poor NHC prediction accuracy and reliability. Recent studies have shown that machine learning can predict the forward speed of vehicles, i.e., virtual odometry(ODO). However, current research mainly discusses predicting virtual NHC and virtual odometry separately, without fully exploring the coupling relationship between the two and the ability of 3D speed to fully constrain the vehicle. Therefore, the study in this paper addresses this issue in depth and proposes an LSTM neural network based on vehicle speed classification for predicting the 3D speed of a vehicle and self-adaptively adjusting its variance domain using 3D speed constraints on the new interest. In order to verify the effectiveness of this paper's method, an experimental test of vehicle-mounted GNSS/SINS combined navigation is carried out. According to the experimental results, the average accuracy of this paper's method in forward velocity prediction is about 0.4 m/s, and the average accuracy in lateral and vertical velocity prediction is about 2 cm/s. In addition, in the case of simulated GNSS signals being out-of-lock for 460 seconds, compared with the inertial derivation results, this paper's method improves the horizontal localization accuracy under the three-dimensional velocity constraint by 99.40%.
  • [1] LI Lan, ZHU Feng, LIU Wanke, et al. GNSS Pseudorange Stochastic Model for Urban Classification Scenes and Its Positioning Performance[J].Geomatics and Information Science of Wuhan University.(李岚,朱锋,刘万科,等.城市分类场景的GNSS伪距随机模型构建及其定位性能分析[J].武汉大学学报(信息科学版).DOI: 10.13203/j.whugis20220598).
    [2] Fu Shuaizhi, Chen Wei, Wu Di, et al.A GNSS/INS Vehicle Integrated Navigation System Based on LSTM-EKF[J].Geomatics And Information Science of Wuhan University, 2023:1-1(傅率智,陈伟,吴迪等.一种基于LSTM-EKF的车载GNSS/INS组合导航系统[J/OL].武汉大学学报(信息科学版),2023:1-11).
    [3] Zhang Hongping,Gao Zhouzheng,Niu Xiaoji,et al.Research on GPS Precise Point with Undifferential and Un-combined Observations[J].Geomatics And Information Science of Wuhan University, 2013,38(12):1396-1399(章红平,高周正,牛小骥,伍岳.GPS非差非组合精密单点定位算法研究[J].武汉大学学报(信息科学版),2013,38(12):1396-1399).
    [4]

    Niu X, Nassar S, El-Sheimy N. An Accurate Land-Vehicle MEMS IMU/GPS Navigation System Using 3D Auxiliary Velocity Updates[J]. Navigation, 2007, 54(3):177-188.

    [5]

    Dissanayake G, Sukkarieh S, Nebot E, et al. The aiding of a low-cost strapdown inertial measurement unit using vehicle model constraints for land vehicle applications[J]. IEEE Transactions on Robotics&Automation, 2001, 17(5):731-747.

    [6]

    Niu X, You L, Quan Z, et al. Observability Analysis of Non-Holonomic Constraints for LandVehicle Navigation Systems[J]. GPS Solutions, 2012, 11(1):80-88.

    [7]

    Liu W, Nong Q, Tao X, et al. OD/SINS adaptive integrated navigation method with nonholonomic constraints[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(1):9-17.

    [8]

    Veneri M, Massaro M. The effect of Ackermann steering on the performance of race cars. Vehicle system dynamics[J]. 2021, 59(6):907-27.

    [9]

    Wen Z, Yang G, Cai Q. An improved SINS/NHC integrated navigation algorithm based on Ackermann turning geometry[J]. Measurement. 2022, 192:110859.

    [10]

    Chen Q, Zhang Q, Niu X. Estimate the pitch and heading mounting angles of the IMU for land vehicular GNSS/INS integrated system[J]. IEEE transactions on intelligent transportation systems. 2020, 22(10):6503-15.

    [11]

    Brossard M, Barrau A, Bonnabel S. AI-IMU dead-reckoning. IEEE Transactions on Intelligent Vehicles[J]. 2020, 5(4):585-95.

    [12]

    Tang H, Niu X, Zhang T, et al. OdoNet:Untethered Speed Aiding for Vehicle Navigation Without Hardware Wheeled Odometer[J]. IEEE Sensors Journal. 2022, 22(12):12197-12208.

    [13]

    Feigl T, Kram S, Woller P, et al. A bidirectional LSTM for estimating dynamic human velocities from a single IMU[C]//In2019 International Conference on Indoor Positioning and Indoor Navigation (IPIN). IEEE, 2019, Pisa, Italy.

    [14]

    Wagstaff B, Kelly J. LSTM-based zero-velocity detection for robust inertial navigation[C]//In2018 International Conference on Indoor Positioning and Indoor Navigation (IPIN). IEEE,2018, Nantes, France.

    [15]

    Saleh K, Hossny M, Nahavandi S. Intent prediction of vulnerable road users from motion trajectories using stacked LSTM network[C]//2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC). IEEE, 2017, Yokohama, Japan,.

    [16]

    Graves A, Mohamed A R, Hinton G. Speech recognition with deep recurrent neural networks[C]//IEEE International Conference on Acoustics. IEEE, 2013,Vancouver, Canada.

    [17] HE Yi, YAO Sheng, CHEN Yi,et al. Spatio-temporal prediction of time-series InSAR Land subsidence based on ConvLSTM neural network[J]. Geomatics and Information Science of Wuhan University.(何毅,姚圣,陈毅,等.ConvLSTM神经网络的时序InSAR地面沉降时空预测[J].武汉大学学报(信息科学版)).
    [18]

    Skog I, Handel P, Nilsson JO, Rantakokko J. Zero-velocity detection-An algorithm evaluation. IEEE transactions on biomedical engineering[J]. 2010 Jul 26;57(11):2657-66.

    [19]

    Zhang X, Zhu F, Tao X, et al. New optimal smoothing scheme for improving relative and absolute accuracy of tightly coupled GNSS/SINS integration[J]. Gps Solutions, 2017, 21(3):861-872.

    [20] Yan Gongmin, Weng Jun. Strapdown inertial navigation algorithm and principle for integrated navigation[M]. Xi'an:Northwestern Polytechnical University Press, 2019(严恭敏,翁浚.捷联惯导算法与组合导航原理[M].2019,西北工业大学出版社).
    [21] YANG Yuanxi. Adaptive navigation and kinematic positioning[M]. Beijing:Surveying and Mapping Press, 2006.(杨元喜.自适应动态导航定位[M].北京:测绘出版社, 2006).
计量
  • 文章访问数:  250
  • HTML全文浏览量:  16
  • PDF下载量:  51
  • 被引次数: 0
出版历程
  • 收稿日期:  2023-12-24
  • 网络出版日期:  2024-01-24

目录

    /

    返回文章
    返回