抗差自适应Kalman滤波及其在GNSS导航中的应用

发布时间：2018-01-29 12:18

本文关键词： Kalman滤波粗差抗差自适应Kalman滤波 GNSS　出处：《河北工程大学学报(自然科学版)》2016年03期 　论文类型：期刊论文

【摘要】：针对Kalman滤波易受粗差影响而导致结果失真的问题,提出一种抗差自适应Kalman滤波方法,该方法结合自适应滤波与抗差Kalman滤波的优点,同时设计自适应因子和抗差因子,采用改进的两段Huber函数与2~3倍的观测噪声中误差来充当抗差因子与粗差判别标准。并对Kalman滤波和抗差自适应滤波(Adaptive Robust Kalman Filtering,ARKF)结果进行比较。车载实验结果表明,ARKF可以有效抵制观测异常对状态估值的影响,同时在系统先验信息不能精确给出的情况下,显著改善了滤波估值的稳定性和可用性。
[Abstract]:Aiming at the problem that Kalman filter is easily affected by gross error, a robust adaptive Kalman filtering method is proposed. This method combines the advantages of adaptive filtering and robust Kalman filtering, and designs the adaptive factor and robust factor at the same time. The improved two-stage Huber function and the observed noise error of 3 times are used as the criterion of robust factor and gross error. The Kalman filter and robust adaptive filter are given. Adaptive Robust Kalman Filtering. The results of vehicle experiments show that ARKF can effectively resist the influence of observed anomalies on the state estimation, and at the same time, when the prior information of the system can not be accurately given. The stability and availability of filter estimation are improved significantly.
【作者单位】：安徽理工大学测绘学院;浙江有色测绘院;
【基金】：国家自然科学基金资助项目(41404004) 安徽省自然科学基金资助项目(1408085QD72) 安徽理工大学科研启动基金资助项目(11152)
【分类号】：P228.4
【正文快照】： Kalman滤波(Kalman Filtering,KF)在GPS导航领域已得到了广泛的应用,只要给出正确的噪声特性,KF将给出滤波状态的最优估计。事实上,动力学模型给出的先验信息不足以表征载体真实的运动状态。这样给出的KF估计值不再是最优估计值,甚至可能导致滤波发散[1-2]。同时,随着导航卫星

【相似文献】