基于双欧拉角的UKF组合导航滤波算法

针对非线性滤波组合导航中四元数无迹估计器(unscented quaternion estimator, USQUE)规范性约束导致的算法计算量大、实时性差等问题, 提出一种基于双欧拉角姿态表示的无迹卡尔曼滤波(dual-Euler unscented Kalman filter, DEUKF)算法。通过正、反欧拉角相互切换, 在正、反欧拉微分方程精华区进行姿态滤波更新, 避免了奇异性, 确保了滤波精度。捷联惯性导航系统/全球定位系统直接式组合导航仿真试验与车载实验结果表明, 相比较于USQUE算法, DEUKF算法计算量小、实时性好, 水平姿态角估计精度与USQUE相当。

Filtering algorithm of UKF integrated navigation based on dual-Euler angles

Aiming at the problem of huge computation load and poor real-time performance caused by the normative constraints of the unscented quaternion estimator (USQUE) in the nonlinear filter for an integrated navigation system, a dual-Euler unscented Kalman filter (DEUKF) based on the dual-Euler attitude representation is proposed. The proposed filter updates attitude in the positive and reverse essence region of the positive and reverse Euler differential equations by switching Euler angle to avoid singularity and ensure filtering accuracy. In the direct integrated navigation of a strapdown inertial navigation system/global positioning system, both the simulation and vehicle test results show that, compared with USQUE algorithm, the computation load of the proposed DEUKF is reduced significantly so that it has good real-time performance while it achieves the equivalent pitch and roll estimation accuracy with USQUE algorithm.