对于低成本SINS／GPS组合导航系统，由于惯性器件的精度较低，通常情况下的SINS误差模型估计不准确，甚至使滤波器发散，为此提出根据姿态四元数的SINS误差估计模型，该模型不需要对初始姿态进行赋值。为在观测噪声未知的情况下估计SINS误差，通过结合序贯处理与Kalman滤波算法，提出一种自适应扩展卡尔曼滤波方法，该方法可以同时进行序贯处理和观测噪声估计。仿真实验结果表明，该方法可以消除过程噪声方差和观测噪声方差不确定造成的影响，提高了SINS／GPS导航系统的性能。%For low‐cost SINS/GPS integrated navigation system ,due to the lower accuracy of inertial devices ,in normal circum‐stances ,SINS error model estimation is not accurate ,the filter divergence problem even emerges .According to attitude quater‐nion ,the SINS error estimation model was proposed which did not require the assignment of initial attitude .To estimate the SINS error in the measurement noise unknown circumstances ,the sequential processing and Kalman filter algorithm were com‐bined into a new method to improve the adaptive ability ,the method can carry out sequential processing and estimate simulta‐neous measurement noise at the same time .The simulation results show that the method can adaptively eliminate the effects brought about by interference noise variance process and observation noise variance uncertainty ,the robustness of SINS/GPS navigation system is improved .
展开▼
