【摘要】 在组合导航系统中，应用常规卡尔曼滤波器（ＫＦ）要求知道系统精确的数学模型和系统噪声与量测噪声的统计特性，才能获得理想的滤波效果，否则，可能产生发散现象。人们越来越倾向于利用自适应滤波（ＡＫＦ）技术来解决发散的问题。针对ＡＫＦ技术的研究现状，本文探讨一种算法结构比较简单、实时性较强、工程上比较实用的在线估计量测噪声统计特性的ＡＫＦ算法。仿真结果表明，这种算法具有较强自适应性，不失为一种实用而有效的滤波方法。更多还原

【Abstract】 There are many practical problems to solve when applying normal KF(Kalman filter) in integrated navigation system, so people are paying more attention to the research of AKF(adaptive Kalman filtering)technique. In view of the present research situation of AKF technique, this paper discusses an AKF algorithm for on line estimation of the statistical feature of measurement noise, which has many advantages such as simple structure, realtime and practicable in engineering. The result of computer simulation shows that the algorithm is useful,effective and adaptive with its high effectiveness and accuracy.更多还原