组合导航抗差自适应卡尔曼滤波

针对组合导航系统的定位精度与稳定性要求不断提高的现状,该文引入一种观测噪声协方差与抗差自适应相结合的Kalman滤波算法。利用新息向量和移动窗口协方差分析法,动态自适应修正观测噪声协方差阵;通过分析基于状态不符值、方差分量的统计量构造的自适应因子所存在的问题,提出一种由预测残差向量构造的自适应因子。仿真结果表明,该方法能够有效抑制观测异常对组合导航定位精度的影响。

Research on robust adaptive Kalman filter for integrated navigation

Aiming at the improvement of the positioning accuracy and stability requirements of integrated navigation system,a Kalman filtering algorithm based on observation noise covariance and robust adaptive is introduced.Using the innovation vector and the moving window covariance analysis,an adaptive factor constructed by the prediction residual vector is proposed by analyzing the problem of the adaptive factor constructed by the statistic of the variance component.Using the innovation vector and the moving window covariance analysis,the observed noise covariance matrix is dynamically modified.By analyzing the problems of adaptive factor constructed by statistical quantity based on state variance and variance component,an adaptive factor constructed by prediction residual vector is proposed.The simulation results show that the proposed method can effectively suppress the influence of observed anomaly on the positioning accuracy of integrated navigation.