针对自主无人水下航行器（AUV）组合导航系统在导航推算时系统模型模糊及测量噪声无法确定导致导航精度下降的问题，提出通过自适应因子调整先验估计误差协方差矩阵的自适应容积卡尔曼滤波（ACKF），以及基于M估计在线调整量测噪声协方差矩阵的鲁棒容积卡尔曼滤波（RCKF），并利用交互式多模型（IMM）将以上优化算法交互融合。结合各个子滤波器的优势，通过设置仿真与实际海试对比实验证明算法的可行性，其中误差降低了29%，均方根误差降低了43%，从而可通过该方法降低AUV导航过程中不同噪声不确定性造成的影响。
Addressing the issues of navigation accuracy degradation in autonomous underwater vehicles (AUVs) due to system model uncertainties and indeterminate measurement noise during navigation calculations in the integrated navigation system, this study proposes two filtering methods. Firstly, an adaptive cubature kalman filter (ACKF) is introduced, which adjusts the prior estimation error covariance matrix through adaptive factors. Secondly, a robust cubature kalman filter (RCKF) is presented, which online tunes the measurement noise covariance matrix based on M-estimation. These optimized algorithms are then interactively fused using an interactive multiple model (IMM) approach. Combining the advantages of each sub-filter, the feasibility of the algorithm was demonstrated through a comparison of simulation and actual sea trials. The results showed a 29% reduction in errors and a 43% decrease in root mean square error. Consequently, this method can mitigate the impact of uncertainties caused by various noises during AUV navigation.
2025,47(5): 37-42 收稿日期：2024-3-20
DOI：10.3404/j.issn.1672-7649.2025.05.006
分类号：U675.73
基金项目：浙江省自然科学基金资助项目（LTGG23E090002）
作者简介：张晓林（1998– ），男，硕士研究生，研究方向为水下机器人的导航算法
参考文献：
[1] 李经. 水下无人作战系统装备现状及发展趋势[J]. 舰船科学技术, 2017, 39(1): 1-5+36.
LI J. Existence and development trend of navy autonomous underwater combat system[J]. Ship Science and Technology, 2017, 39(1): 1-5+36.
[2] 卢道华, 宋世磊, 王佳, 等. SINS/DVL水下组合导航技术发展综述[J]. 控制理论与应用, 2022, 39(7): 1159-1170.
LU D H, SONG S L, WANG J, et al. Review of the development of SINS/DVL underwater integrated navigation technology[J]. Control Theory & Application, 2022, 39(7): 1159-1170.
[3] GAO B, HU G, ZHONG Y, 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.
[4] 王迪, 王冰, 黄浩乾, 等. 复杂水下环境中自适应粒子群优化的SINS/DVL组合导航方法[J]. 中国惯性技术学报, 2023, 31(10): 1023-1029+1036.
WANG D, WANG B, HUANG H Q, et al. SINS/DVL integrated navigation method based on adaptive particle swarm optimization in complex underwater environments[J]. Journal of Chinese Inertial Technology, 2023, 31(10): 1023-1029+1036.
[5] BUCCI A, FRANCHI M, RIDOLFI A, et al. Evaluation of UKF-based fusion strategies for autonomous underwater vehicles multisensor navigation[J]. IEEE Journal of Oceanic Engineering, 2023, 48(1): 1-26.
[6] ZHANG X, HE B, GAO S, et al. Multiple model AUV navigation methodology with adaptivity and robustness[J]. Ocean Engineering, 2022, 254: 111258.
[7] FRASER C T, ULRICH S. Adaptive extended Kalman filtering strategies for spacecraft formation relative navigation[J]. Acta Astronautica, 2021, 178: 700-721.
[8] ODRY Á, KECSKES I, SARCEVIC P, et al. A novel fuzzy-adaptive extended kalman filter for real-time attitude estimation of mobile robots[J]. Sensors, 2020, 20(3): 803.
[9] HU G, NI L, GAO B, et al. Model predictive based unscented kalman filter for hypersonic vehicle navigation with INS/GNSS Integration[J]. IEEE Access, 2020, 8: 4814-4823.
[10] ARASARATNAM I, HAYKIN S. Cubature Kalman filters[J]. IEEE Transactions on Automatic Control, 2009, 54(6): 1254-1269.
[11] MENG D, MIAO L, SHAO H, et al. A seventh-degree cubature Kalman filter[J]. Asian Journal of Control, 2018, 20(1): 250-262.
[12] 葛磊. 容积卡尔曼滤波算法研究及其在导航中的应用[D]. 哈尔滨: 哈尔滨工程大学, 2013.
[13] 张荣辉, 贾宏光, 陈涛, 等. 基于四元数法的捷联式惯性导航系统的姿态解算[J]. 光学精密工程, 2008(10): 1963-1970.
ZHANG R H, JIA H G, CHEN T, et al. Attitude calculation of strapdown inertial navigation system based on quaternion method[J]. Optics and Precision Engineering, 2008(10): 1963-1970.
[14] 肖春亮, 张忠民. 基于改进容积卡尔曼滤波器的超宽带/惯性导航系统室内融合定位技术[J]. 应用科技, 2023, 50(6): 69-75.
XIAO C L, ZHANG Z M. Indoor fusion positioning technology of ultra-wideband/inertial navigation system based on improved cubature kalman filter[J]. Applied Science and Technology, 2023, 50(6): 69-75.
