基于Allan方差的GNSS/SINS组合导航改进自适应滤波算法

Improved Adaptive Filtering Algorithm for GNSS/SINS Integrated Navigation Based on Allan Variance

摘要:

GNSS信号易受到干扰，进而导致GNSS/SINS组合导航系统中测量噪声方差突变，经典的基于Allan方差法的自适应卡尔曼滤波（ALAKF）是估计未知测量噪声方差的有效方法，然而存在检测精度低的问题，为此提出了一种基于Allan方差法的改进自适应卡尔曼滤波（AL_IAKF）。首先，针对测量噪声方差突变的开始时刻及结束时刻难以检测的问题，构建了基于正常测量噪声方差的残差卡方检测准则。然后，针对ALAKF无法准确跟踪测量噪声方差突变的“上升沿”及“下降沿”的问题，构造可变的遗忘因子模型以动态调整AL_IAKF中的遗忘因子。最后，进行了基于ALAKF和AL_IAKF的组合导航对比实验。实验结果表明，相较于ALAKF，AL_IAKF可明显提高对突变和缓变的测量噪声方差估计精度，进而提高了组合导航系统的滤波精度。

Abstract:

s: GNSS signals are susceptible to interference, which can lead to sudden changes in the measurement noise variance in the GNSS/SINS integrated navigation system. The classic adaptive Kalman filtering based on Allan variance method (ALAKF) is an effective way to estimate the unknown measurement noise variance. However, it has the problem of low detection accuracy. Therefore, an improved adaptive Kalman filtering based on Allan variance method (AL_IAKF) is proposed. Firstly, to address the difficulty in detecting the beginning and ending times of the variance mutation of the measurement noise, a residual chi-square detection criterion based on the normal measurement noise variance was constructed. Then, to solve the problem that ALAKF cannot accurately track the "rising edge" and "falling edge" of the measurement noise variance mutation, a variable forgetting factor model was constructed to dynamically adjust the forgetting factor in AL_IAKF. Finally, a comparative navigation experiment based on ALAKF and AL_IAKF was conducted. The experimental results show that, compared with ALAKF, AL_IAKF can significantly improve the estimation accuracy of the measurement noise variance for sudden and gradual changes, thereby enhancing the filtering accuracy of the combined navigation system.