摘要
从卡尔曼滤波算法的稳定性出发，分析了卡尔曼滤波算法和Sage-Husa滤波算法的缺点，改进了Sage-Husa滤波算法，改进后的算法保证了系统噪声估计方差阵和测量噪声估计方差阵的半正定性和正定性，消除了滤波发散现象。在INS／GPS组合导航系统中进行仿真，仿真结果表明改进后的算法可以有效提高INS／GPS组合导航系统的稳定性．
Considering the stability of Kalman filter, the paper discussed the disadvantage of Sage-Husa filter and brought forward a new improved Sage-Husa filter which can ensure the half positive of the matrix of system noisy estimation and the positive the matrix of measure noisy estimation, hence resulting in no filtering riveting. The simulation shows that the improved algorithm can improve the fault-tolerant and computing capability performance of GPS/INS integrated navigation system.
出处
《导航》
2005年第3期107-112,共6页
