论文部分内容阅读
针对普通UKF(无迹卡尔曼滤波)测量更新方法的非线性近似精度相对较低,导致目标跟踪滤波精度和稳定性较低的问题,在单星对空间目标的天基仅测角跟踪滤波过程中,提出一种基于迭代测量更新方法的IUKF(迭代UKF)算法。通过在测量更新过程中提高非线性系统状态估计的近似精度,进而提高目标跟踪滤波精度,并引入具有全局收敛性的阻尼Gauss-Newton(高斯-牛顿)法来改进IUKF的数值稳定性。理论分析与实验结果表明,该方法不仅避免了求解雅可比矩阵和Hessian矩阵,而且具有较高的滤波精度和数值稳定性。
The nonlinear approximation accuracy of the conventional UKF (Unscented Kalman Filter) measurement and updating method is relatively low, which leads to the problem of low accuracy and stability of the target tracking filtering. In the single-star space-only space-based tracking algorithm , An IUKF (Iterative UKF) algorithm based on iterative measurement updating method is proposed. By improving the approximate accuracy of the nonlinear system state estimation during the measurement and updating, the accuracy of the target tracking filter is improved, and the damped Gauss-Newton method with global convergence is introduced to improve the numerical stability of IUKF. Theoretical analysis and experimental results show that this method not only avoids solving Jacobi matrix and Hessian matrix, but also has higher filtering accuracy and numerical stability.