降维CKF算法在大失准角传递对准中的应用

惯性导航具备完全自主、高度隐蔽、数据频率高等优点，在机载精确制导武器中得到广泛应用。由于空中传递对准可能处于恶劣条件下，初始失准角较大，传统的线性传递对准模型不能反映系统的真实情况，线性滤波算法也将导致很大的对准误差，为实现精确对准，需要应用非线性模型和非线性滤波算法。而非线性滤波算法如无迹卡尔曼滤波(UKF)、容积卡尔曼滤波(CKF)等，在高维情况下，计算量很大，在弹载计算机计算资源受限的情况下，如何降低滤波算法的计算量是一个重要问题。本文将降维容积卡尔曼滤波算法应用在非线性传递对准模型中，将容积卡尔曼滤波的采样点由30个减少为6个，大幅减少了所需计算量。

Dimension reduced CKF algorithm for transfer alignment with large misalignment angle

Inertial navigation is completely autonomous, highly covert and has a high data frequency, which is widely used in airborne guided weapons. As the condition for transfer alignment may be bad during the flight, and the misalignment angle may be large, so that traditional linear model cannot accurately reflect the real situation of the system. To achieve a high alignment accuracy, nonlinear model and filter are utilized. Because nonlinear filter has a large calculation quantity and the computing power of on-board computer is limited, it is very important to reduce the calculated quantity of the filter. A dimension reduced Cubature Kalman Filter(CKF) algorithm is adopted in nonlinear transfer alignment model, and the sampling points are reduced from 30 to 6, significantly reducing the calculation quantity.