考虑非高斯相关噪声的视线角速率提取算法

针对捷联导引头敏感信息存在非高斯噪声且状态噪声与观测噪声相关的问题，提出了修正球面坐标系下带有相关噪声解耦的扩展椭球集员滤波（ESMF）算法。首先，通过对弹目相对距离及其变化率的规范化处理，得到修正球面坐标系下的视线角速率提取模型；然后，利用相关噪声解耦方法去除观测方程中耦合的状态噪声项，基于泰勒级数展开推导了相关噪声解耦后的模型线性化表达式，得到包含线性化误差的虚拟噪声椭球集合，并根据最小迹和最小化尺度因子上界的优化方法得到状态更新与观测更新椭球，形成非高斯相关噪声解耦的扩展椭球集员滤波算法。仿真结果表明，所提算法能够在考虑非高斯相关噪声的情况下，实现捷联导引头视线角速率的高精度提取。

Extraction algorithm for line of sight angular rate in non-Gaussian correlative noise environment

To extract the Line of Sight (LOS) angular rate of a strapdown seeker with non-Gaussian noises and the correlation between the state noises and the observation noises, an Extended Set Membership Filter (ESMF) with correlated noises decoupling in a modified spherical coordinate system is proposed. An LOS angular rate extraction model in the modified spherical coordinate system is first established by normalizing the relative distance and the rate of change between the projectile and the target. The correlated noises decoupling method is then used to remove the correlated noises coupled in the observation equation, and the Taylor series expansion employed to derive the linearization expression of the decorrelation model to obtain a set of virtual noise ellipsoids containing the linearization error. Moreover, the ellipsoid for state prediction and observation update is optimized by means of minimum trace and minimum upper bound of the scale factor. The simulation results show that the ESMF can achieve good estimation precision with non-Gaussian correlated noises and realize high precision LOS angular rate extraction of the strapdown seeker.