GNSS/SINS组合导航自适应最大相关熵滤波算法研究

Adaptive Maximum Correntropy Filtering Algorithm for GNSS/SINS Integrated Navigation

ES评分 0

DOI 10.14075/j.jgg.2024.11.528
刊名
Journal of Geodesy and Geodynamics
年,卷(期) 2025, 45(10)
作者
作者单位

1 山东外事职业大学人工智能学院,山东省威海市长江路788号, 264500;
2 海军航空大学,山东省烟台市, 264001

摘要
为提高复杂环境下最大相关熵滤波(MCKF)方法的滤波性能,提出一种GNSS/SINS组合导航系统的自适应最大相关熵滤波算法(AMCKF)。首先,根据当前和过去l-1个时刻的滤波故障检测函数值与检测门限的关系,形成0和1两种映射关系;然后,以映射值为基础构成一个长度为l的窗,进而提出一种核宽度自适应调整算法;最后,将该核宽度自适应调整算法应用于MCKF中,进而构成组合导航系统的AMCKF算法。在测量高斯白噪声和重尾脉冲噪声环境下搭建GNSS/SINS组合导航系统实验,结果表明,AMCKF可以根据不同的测量噪声环境自适应调整核带宽值,进而提高组合导航系统的滤波精度;相对于KF和MCKF,AMCKF可分别提高位置精度约26.5%和16.4%,提高速度精度约15.5%和6.4%
Abstract
To enhance the filtering performance of the maximum correntropy Kalman filter(MCKF) method in complex environments, an adaptive maximum correntropy Kalman filter algorithm(AMCKF) for GNSS/SINS integrated navigation systems is proposed. Firstly, based on the relationship between the filtering fault detection function values at the current and the past (l-1) time steps and the detection threshold, two mapping relationships of 0 and 1 are formed. Then, a window of length l is constructed using the mapped values, and a kernel width adaptive adjustment algorithm is proposed. Finally, this kernel width adaptive adjustment algorithm is applied to MCKF to form the AMCKF algorithm for the integrated navigation system. Experiments are conducted on a GNSS/SINS integrated navigation system under measurement environments with Gaussian white noise and heavy-tailed impulsive noise. The results indicate that AMCKF can adaptively adjust the kernel width value according to different measurement noise environments, thereby improving the filtering accuracy of the integrated navigation system. Compared with KF and MCKF, AMCKF can improve position accuracy by approximately 26.5% and 16.4%, respectively, and improve velocity accuracy by approximately 15.5% and 6.4%, respectively.
关键词
组合导航; 最大相关熵准则; 核带宽; 自适应滤波
KeyWord
integrated navigation; maximum correntropy criterion; kernel bandwidth; adaptive filter
基金项目
页码 1043-1048
  • 参考文献
  • 相关文献
  • 引用本文

林雪原, 潘新龙, 孙玉霞. GNSS/SINS组合导航自适应最大相关熵滤波算法研究 [J]. 大地测量与地球动力学. 2025; 45; (10). 1043 - 1048.

  • 文献评论

相关学者

相关机构