H∞滤波及其在INS CNS组合导航系统中的应用

在惯导/星光（INS/CNS）组合导航中，由于集中卡尔曼滤波存在对系统模型和噪声统计特性依赖性等缺点，将H∞滤波方法理论应用其中。给出了H∞滤波的滤波方程，定性讨论了kalman滤波和H∞滤波的关系，通过在INS/CNS组合导航系统中的实际应用进一步从精度和鲁棒性对两种滤波算法进行了比较。仿真结果表明，当噪声为有色噪声或统计特性不确定时，H∞滤波优于kalman滤波，说明了该算法的有效性和可行性。

Abstract:

 As the special dependence on accuracy of the system model and statistic character of noises of kalman filter when it is used in INS/CNS integrated navigation system， H∞ filter is applied. Equations of H∞ filtering algorithm is presented and the relationship between H∞ filtering algorithm and traditional kalman filter is discussed. Then in a INS/CNS integrated navigation system， the accuracy and robustness of the two algorithm are compared. The results of simulation demonstrate H∞ filtering is better than kalman filtering when the colored noises or the uncertainty of the model is existed， it proves the effectiveness and feasibility of the presented method.