为了降低非线性对准的计算量而不损失对准精度,针对容积卡尔曼滤波(CKF)采样点数与状态维数成正比、计算量较大的问题,提出了基于简化CKF/降维CKF混合滤波的非线性对准方法。利用大失准角模型和基于线性观测方程的简化CKF算法进行水平对准;使用大方位失准角模型和降维CKF完成精对准。仿真结果表明,该方法摆脱了CKF算法的“维数灾难”和降维CKF对准应用条件限制,能够完成任意失准角下的初始对准并获得较高对准精度,具有重要的工程应用价值。

In order to reduce calculation amount and keep alignment precision of nonlinear alignment, the problems that the sample points are directly proportional to state dimension and the calculation amount is large in cubature Kalman filter (CKF), a new alignment algorithm with mixed filter based on simplified CKF (SCKF) and reduced dimension CKF(RDCKF) proposed. The level alignment finished by a large misalignment angle model and SCKF without coarse alignment; the fine alignment fulfilled by a large azimuth misalignment angle model and RDCKF based on the level alignment. The simulation result shows that this way two disadvantages that CKF's "dimension problem" and RDCKF's application limitation. It is available on any misalignment angle and has higher precision, and with important engineering application value.