基于鲁棒高阶容积滤波的无人机相对导航状态估计方法

Unmanned aerial vehicle relative navigation method based on robust high degree cubature filtering

由于无人机相对导航系统具有非线性强、噪声非高斯的特点,传统的基于卡尔曼滤波算法设计的相对导航滤波器存在估计失准甚至发散的问题。考虑到高阶容积卡尔曼滤波和最大熵滤波算法分别在解决非线性问题和非高斯问题时的优势,利用最大熵滤波的量测更新方法对高阶容积卡尔曼滤波的测量更新方程进行了改进,将传统的量测更新问题转换成了线性衰退的求解问题,避免了对测量噪声进行高斯假设,同时解决了系统非线性和量测噪声非高斯的问题。进行了相应的数学仿真,仿真结果表明:所提算法的估计精度超过了高阶容积卡尔曼滤波和最大熵滤波算法的,验证了算法的有效性。

Due to the strong-nonlinearity and non-Gaussianity of the UAV（ unmanned aerial vehicle） relative navigation system,the accuracy of the traditional relative navigation filter,which is designed based on the Kalman filtering algorithm,decreases or even diverge. In view of the advantages of HCKF（ high degree cubature Kalman filter） and MCKF（ maximum correntropy Kalman filter） in coping with nonlinear problems and non-Gaussian problems,respectively,the measurement update equation was modified by the measurement update method of MCKF,and the traditional measurement update problem was recast as a linear regression problem. In addition,the Gaussian assumption of the measurement noise was avoided,the system nonlinearity and measurement noise non-Gaussianity were solved at the same time. The simulation was conducted,and the simulation results indicated that RHCF is superior to HCKF and MCKF. Hence,the effectiveness of the proposed algorithm is verified.