引用本文:吕东辉,王炯琦,熊凯,侯博文,何章鸣.适用处理非高斯观测噪声的强跟踪卡尔曼滤波器[J].控制理论与应用,2019,36(12):1997~2004.[点击复制]
LV Dong-hui,WANG Jiong-qi,XIONG Kai,HOU Bo-wen,HE Zhang-ming.Strong tracking Kalman filter for non-Gaussian observation[J].Control Theory and Technology,2019,36(12):1997~2004.[点击复制]
适用处理非高斯观测噪声的强跟踪卡尔曼滤波器
Strong tracking Kalman filter for non-Gaussian observation
摘要点击 2875  全文点击 798  投稿时间：2019-07-07  修订日期：2019-09-14
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2019.90535
  2019,36(12):1997-2004
中文关键词  卡尔曼滤波器，强跟踪滤波器，非高斯观测噪声，滤波性能
英文关键词  Kalman filter, strong tracking filter, non-Gaussian observation noise, filter performance
基金项目  国家自然科学基金,国家杰出青年科学基金,省自然科学基金,其它
作者单位E-mail
吕东辉 国防科技大学文理学院 290932317@qq.com 
王炯琦* 国防科技大学文理学院 wjq_gfkd@163.com 
熊凯 北京控制工程研究所  
侯博文 国防科技大学文理学院  
何章鸣 国防科技大学文理学院  
中文摘要
      在高斯噪声条件下, 卡尔曼滤波器(KF)能够获得系统状态的一致最小方差线性无偏估计. 但当噪声非高斯, KF性能将严重下降. 观测噪声非高斯现象在深空探测自主导航中经常遇到, 然而现有模型可能存在着精度不高、稳定性不强或者计算复杂度较高的缺点. 针对这种现状, 本文在传统强跟踪卡尔曼滤波器(STKF)中新息正交原则的基础上, 推导了适用处理非高斯观测噪声的强跟踪卡尔曼滤波器(STKFNO), 并将其嵌入到无迹卡尔曼滤波(UKF)框架下形成适用处理非线性系统非高斯观测噪声的强跟踪无迹卡尔曼滤波器(STUKFNO). 所提出的算法被应用到深空光学自主导航系统中, 仿真结果表明所提出的算法能够较好地应对观测噪声的非高斯性.
英文摘要
      Under Gaussian noise, Kalman Filter (KF) can obtain the uniformly minimum variance linear unbiased estimation of system state. However, when the noise is non-Gaussian, the performance of KF will degrade seriously. Non-Gaussian phenomena of observation noise are often encountered in autonomous navigation of deep space exploration. However, the existing models may have drawbacks such as low accuracy, low stability or high computational complexity. In view of this situation, based on the orthogonal principle of innovation in traditional strong tracking Kalman filter (STKF), strong tracking Kalman filter for non-Gaussian observation (STKFNO) which is applicable to processing non-Gaussian observation noise is derived. By embedding STKFNO into the framework of unscented Kalman filter(UKF), strong tracking unscented Kalman filter for non-Gaussian observation (STUKFNO) suitable for dealing with non-Gaussian noise of non-linear systems is also established. The proposed algorithm is applied to a deep space optical autonomous navigation system. The simulation results demonstrate that the proposed algorithm is effective in disposing of non-Gaussian observation noise.