七阶正交容积卡尔曼滤波算法

A seventh-degree cubature quadrature Kalman filter

在高斯滤波框架下,阶次越高,近似精度越高。为提高滤波精度,通过提高阶次,提出了七阶正交容积卡尔曼滤波(CQKF)算法。在传统CQKF算法的基础上,该算法扩展了线性积分的近似阶次,提出了七阶球面积分的确定性采样方法;进而扩展了球-半径准则,提高了滤波估计精度。飞行器目标跟踪的仿真实验证明了该算法的有效性,证明了七阶CQKF比五阶CQKF、三阶容积卡尔曼滤波器(CKF)和无迹卡尔曼滤波器(UKF)有更高的滤波精度。

In the Gaussian filter frame,the higher the order,the higher the accuracy of the approximation.To improve the filtering accuracy,a seventh-degree Cubature Quadrature Kalman Filter（CQKF）algorithm is proposed by improving the degree.Based on the traditional CQKF,the algorithm extends the approximate degree of linear integrals,and proposes a deterministic sampling method for the seven-degree spherical integral.Then the spherical-radial rule is extended to improve the accuracy of the filter.The simulation results of the aircraft target tracking demonstrate the effectiveness of the algorithm.It is proved that the seventh-degree CQKF is more accurate than the fifth-CQKF,third-degree Cubature Kalman Filter（CKF）and Unscented Kalman Filter（UKF）.