论文部分内容阅读
为了提高惯性/卫星深组合导航系统的滤波性能,在抗差自适应滤波算法的基础上,研究了一种优化抗差自适应滤波算法。该算法通过比较实际预测残差协方差矩阵和理论协方差阵的差值来生成自适应因子,从而优化抗差自适应滤波。将所研究的算法应用于惯性/卫星深组合导航系统,在高动态环境下进行仿真验证,并与常规卡尔曼滤波、抗差自适应滤波进行比较。结果表明,优化算法能有效地控制观测异常和动态模型异常对状态参数估值的影响,所得组合导航位置误差和速度误差明显减小,提高了组合导航系统的滤波精度。
In order to improve the filtering performance of inertial / satellite deep integrated navigation system, an adaptive robust adaptive filtering algorithm is studied on the basis of robust adaptive filtering algorithm. The algorithm generates the adaptive factor by comparing the difference between the actual predicted residual covariance matrix and the theoretical covariance matrix, so as to optimize the robust adaptive filter. The proposed algorithm is applied to the inertial / satellite deep integrated navigation system, and the simulation is carried out in high dynamic environment. Compared with the conventional Kalman filter and the robust adaptive filter, the proposed algorithm is validated. The results show that the optimization algorithm can effectively control the influence of abnormal observation and dynamic model anomaly on the estimation of state parameters. The resulting integrated navigation position error and velocity error decrease significantly, which improves the filtering accuracy of integrated navigation system.