搜索

x

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

逼近积分点数下限的五阶容积卡尔曼滤波定轨算法

李兆铭 杨文革 丁丹 廖育荣

引用本文:
Citation:

逼近积分点数下限的五阶容积卡尔曼滤波定轨算法

李兆铭, 杨文革, 丁丹, 廖育荣

A novel algorithm of fifth-degree cubature Kalman filter for orbit determination at the lower bound approaching to the number of cubature points

Li Zhao-Ming, Yang Wen-Ge, Ding Dan, Liao Yu-Rong
PDF
导出引用
  • 为了在保持滤波定轨精度不变的条件下提高定轨计算的实时性,提出一种新的逼近积分点个数下限的五阶容积卡尔曼滤波定轨算法.首先,采用一种数值容积准则对非线性函数的高斯加权积分进行近似,该准则所需的积分点个数仅比五阶代数精度容积准则积分点个数的理论下限多一个积分点,并在贝叶斯滤波算法框架下推导出本文算法的更新步骤.然后,给出实时定轨所需的状态方程和量测方程,在状态方程中考虑了J2项引力摄动和大气阻力摄动,在量测方程中利用坐标系转换推导了轨道状态与测量元素之间的非线性关系.仿真实验结果表明,本文所提算法在定轨精度方面与已有的五阶滤波算法相当,但所需的积分点个数最少,计算实时性最高,从而验证了本文算法的有效性.
    With more satellites launched into orbits during recent years, monitoring and cataloging of satellites play an important role in improving the utilization rate of space resource and alleviating the pressure of orbit resource. Groundbased radar, a kind of sensor in space surveillance system, does not consider the influences of the weather and other special circumstances. And it is a key technology in space target tracking by using the measurement data for real-time orbit determination. Due to the influence of orbital perturbation, the satellite orbital dynamic model is a nonlinear system. The optimal estimation of the orbital state can be achieved by means of nonlinear filtering based on the measured ranging, velocity and angle data with measurement noise, which is the essence of real time orbit determination and has important research value. The extended Kalman filter (EKF) and unscented Kalman filter (UKF) are most widely used nonlinear Kalman filters. However, the first-order Taylor expansion of nonlinear function in EKF degrades the filtering accuracy. And the weight value in UKF might be negative for the high-dimensional system, which may directly affect the filtering stability. As an important method in nonlinear filtering, cubature Kalman filter (CKF) has better accuracy and stability than UKF. However, CKF only has third-degree filtering accuracy. In order to improve the filtering accuracy further, some fifth-degree cubature Kalman filters are proposed, mainly including the fifth-degree cubature Kalman filter and the fifth-degree spherical simplex-radial cubature Kalman filter. The optimality of the radial integral cannot be guaranteed by using the moment matching method in these fifth-degree filters, so a high-degree cubature quadrature Kalman filter (HDCQKF) is proposed. The radial integral is calculated using the high-degree Gauss-Laguerre formula in HDCQKF. However, the aforementioned filtering algorithm leads to an increase in the number of cubature points, thereby improving the accuracy, and the number of cubature points increases polynomially with the increase of system dimension. Once the algorithm is applied to a high-dimensional system, or the processor has a relatively poor performance, it may impose a heavier computing burden, thus the real-time performance decreases. Therefore, it is necessary to study how to reduce the computational complexity of the fifth-degree filtering algorithm. In order to improve the real-time performance of orbit determination on condition that the accuracy of orbit determination is kept, a novel fifth-degree cubature Kalman filter for orbit determination is proposed at the lower bound approaching to the number of cubature points. The key problem in the nonlinear Kalman filter is to calculate the multidimensional integral in the form of nonlinear functionGaussian probability density function, and the integral is approximated using a fifth-degree numerical cubature rule, in which the number of cubature points required is only one more than the theoretical lower bound. The abovementioned cubature rule is embedded into the nonlinear Kalman filtering framework, from which the update steps of the novel cubature Kalman filter are derived. Then, the equations of state and measurement for real-time orbit determination are obtained. The J2 perturbation and atmospheric drag perturbation are taken into account in the state equation, and the coordinate transformation is used to derive the nonlinear relationship between the orbital state and measurement element. The simulation results show that the proposed fifth-degree cubature Kalman filter can achieve a higher filtering accuracy than the CKF and the same accuracy as the existing fifth-degree filters, but has the fewest cubature points and the best real-time performance, which proves the effectiveness of the proposed algorithm.
      通信作者: 杨文革, wengeyang_3@163.com
    • 基金项目: 国家高技术研究发展计划(批准号:2015AA7026085)资助的课题.
      Corresponding author: Yang Wen-Ge, wengeyang_3@163.com
    • Funds: Project supported by the National High Technology Research and Development Program of China (Grant No.2015AA7026085).
    [1]

    Ning X, Ye C M, Yang J, Shen B 2014 Chin. J. Radio 29 27 (in Chinese) [宁夏, 叶春茂, 杨健, 沈彬 2014 电波科学学报 29 27]

    [2]

    Abhinoy K S, Shovan B 2014 IEEE International Symposium on Signal Processing and Information Technology Noida, India, December 15-17, 2014 p114

    [3]

    Arasaratnam I, Haykin S 2009 IEEE Trans. Autom. Control 54 1254

    [4]

    Jafar Z, Ehsan S 2015 IET Sci. Meas. Technol. 9 294

    [5]

    Julier S J, Uhlmann J K, Whyte H F D 2000 IEEE Trans. Autom. Control 45 477

    [6]

    Xiong K, Zhang H Y, Chan C W 2006 Automatica 42 261

    [7]

    Zhang L J, Yang H B, Lu H P, Zhang S F, Cai H, Qian S 2014 Acta Astronaut. 105 254

    [8]

    Chen J G, Wang N, Ma L L, Xu B G 2015 IET Radar Sonar Navig. 9 324

    [9]

    Lu Z Y, Wang D M, Wang J H, Wang Y 2015 Acta Phys. Sin. 64 150502 (in Chinese) [逯志宇, 王大鸣, 王建辉, 王跃 2015 物理学报 64 150502]

    [10]

    Wu H, Chen S X, Yang B F, Chen K 2015 Acta Phys. Sin. 64 218401 (in Chinese) [吴昊, 陈树新, 杨宾峰, 陈坤 2015 物理学报 64 218401]

    [11]

    Jia B, Xin M, Cheng Y 2012 Automatica 49 510

    [12]

    Jia B, Xin M, Cheng Y 2012 IEEE Conference on Decision and Control Maui Hawaii, USA, December 10-13, 2012 p4095

    [13]

    Huang X Y, Tang X Q, Wu M 2015 Syst. Eng. Electron. 37 633 (in Chinese) [黄湘远, 汤霞清, 武萌 2015 系统工程与电子技术 37 633]

    [14]

    Zhang X C 2014 Circuits Syst. Signal Process 33 1799

    [15]

    Zhang L, Cui N G, Wang X G, Yang F, Lu B G 2015 Acta Aeronaut. Astronaut. Sin. 36 3885 (in Chinese) [张龙, 崔乃刚, 王小刚, 杨峰, 卢宝刚 2015 航空学报 36 3885]

    [16]

    Zhang W J, Wang S Y, Feng Y L, Feng J C 2016 Acta Phys. Sin. 65 088401 (in Chinese) [张文杰, 王世元, 冯亚丽, 冯久超 2016 物理学报 65 088401]

    [17]

    Zhao L Q, Chen K Y, Wang J L, Yu T 2016 Control Decis. 31 1080 (in Chinese) [赵利强, 陈坤云, 王建林, 于涛 2016 控制与决策 31 1080]

    [18]

    Wang S Y, Feng J C, Tse C K 2014 IEEE Signal Process. Lett. 21 43

    [19]

    Singh A K, Bhaumik S 2015 Int. J. Control Autom. Syst. 13 1097

    [20]

    Lu J, Darmofal D L 2004 SIAM J. Sci. Comput. 26 613

  • [1]

    Ning X, Ye C M, Yang J, Shen B 2014 Chin. J. Radio 29 27 (in Chinese) [宁夏, 叶春茂, 杨健, 沈彬 2014 电波科学学报 29 27]

    [2]

    Abhinoy K S, Shovan B 2014 IEEE International Symposium on Signal Processing and Information Technology Noida, India, December 15-17, 2014 p114

    [3]

    Arasaratnam I, Haykin S 2009 IEEE Trans. Autom. Control 54 1254

    [4]

    Jafar Z, Ehsan S 2015 IET Sci. Meas. Technol. 9 294

    [5]

    Julier S J, Uhlmann J K, Whyte H F D 2000 IEEE Trans. Autom. Control 45 477

    [6]

    Xiong K, Zhang H Y, Chan C W 2006 Automatica 42 261

    [7]

    Zhang L J, Yang H B, Lu H P, Zhang S F, Cai H, Qian S 2014 Acta Astronaut. 105 254

    [8]

    Chen J G, Wang N, Ma L L, Xu B G 2015 IET Radar Sonar Navig. 9 324

    [9]

    Lu Z Y, Wang D M, Wang J H, Wang Y 2015 Acta Phys. Sin. 64 150502 (in Chinese) [逯志宇, 王大鸣, 王建辉, 王跃 2015 物理学报 64 150502]

    [10]

    Wu H, Chen S X, Yang B F, Chen K 2015 Acta Phys. Sin. 64 218401 (in Chinese) [吴昊, 陈树新, 杨宾峰, 陈坤 2015 物理学报 64 218401]

    [11]

    Jia B, Xin M, Cheng Y 2012 Automatica 49 510

    [12]

    Jia B, Xin M, Cheng Y 2012 IEEE Conference on Decision and Control Maui Hawaii, USA, December 10-13, 2012 p4095

    [13]

    Huang X Y, Tang X Q, Wu M 2015 Syst. Eng. Electron. 37 633 (in Chinese) [黄湘远, 汤霞清, 武萌 2015 系统工程与电子技术 37 633]

    [14]

    Zhang X C 2014 Circuits Syst. Signal Process 33 1799

    [15]

    Zhang L, Cui N G, Wang X G, Yang F, Lu B G 2015 Acta Aeronaut. Astronaut. Sin. 36 3885 (in Chinese) [张龙, 崔乃刚, 王小刚, 杨峰, 卢宝刚 2015 航空学报 36 3885]

    [16]

    Zhang W J, Wang S Y, Feng Y L, Feng J C 2016 Acta Phys. Sin. 65 088401 (in Chinese) [张文杰, 王世元, 冯亚丽, 冯久超 2016 物理学报 65 088401]

    [17]

    Zhao L Q, Chen K Y, Wang J L, Yu T 2016 Control Decis. 31 1080 (in Chinese) [赵利强, 陈坤云, 王建林, 于涛 2016 控制与决策 31 1080]

    [18]

    Wang S Y, Feng J C, Tse C K 2014 IEEE Signal Process. Lett. 21 43

    [19]

    Singh A K, Bhaumik S 2015 Int. J. Control Autom. Syst. 13 1097

    [20]

    Lu J, Darmofal D L 2004 SIAM J. Sci. Comput. 26 613

  • [1] 刘希望, 张宏丹, 贲帅, 杨士栋, 任鑫, 宋晓红, 杨玮枫. 费曼路径积分强场动力学计算方法. 物理学报, 2023, 72(19): 198701. doi: 10.7498/aps.72.20230451
    [2] 朱栋, 徐晗, 周寅, 吴彬, 程冰, 王凯楠, 陈佩军, 高世腾, 翁堪兴, 王河林, 彭树萍, 乔中坤, 王肖隆, 林强. 基于扩展卡尔曼滤波算法的船载绝对重力测量数据处理. 物理学报, 2022, 71(13): 133702. doi: 10.7498/aps.71.20220071
    [3] 丁明松, 江涛, 刘庆宗, 董维中, 高铁锁, 傅杨奥骁. 基于电流积分计算磁矢量势修正的低磁雷诺数方法. 物理学报, 2020, 69(13): 134702. doi: 10.7498/aps.69.20200091
    [4] 冯玲, 纪婉妮. 随机波动率费曼路径积分股指期权定价. 物理学报, 2019, 68(20): 203101. doi: 10.7498/aps.68.20190714
    [5] 易洪, 李松, 马跃, 黄科, 周辉, 史光远. 基于足印探测的激光测高仪在轨标定. 物理学报, 2017, 66(13): 134206. doi: 10.7498/aps.66.134206
    [6] 何秋燕, 袁晓. Carlson与任意阶分数微积分算子的有理逼近. 物理学报, 2016, 65(16): 160202. doi: 10.7498/aps.65.160202
    [7] 张文杰, 王世元, 冯亚丽, 冯久超. 基于Huber的高阶容积卡尔曼跟踪算法. 物理学报, 2016, 65(8): 088401. doi: 10.7498/aps.65.088401
    [8] 吴昊, 陈树新, 杨宾峰, 陈坤. 基于广义M估计的鲁棒容积卡尔曼滤波目标跟踪算法. 物理学报, 2015, 64(21): 218401. doi: 10.7498/aps.64.218401
    [9] 逯志宇, 王大鸣, 王建辉, 王跃. 基于时频差的正交容积卡尔曼滤波跟踪算法. 物理学报, 2015, 64(15): 150502. doi: 10.7498/aps.64.150502
    [10] 李一丁, 张鹏飞, 张辉, 于淼. 弯轨Čerenkov辐射计算中的稳相法. 物理学报, 2013, 62(10): 104103. doi: 10.7498/aps.62.104103
    [11] 徐峰, 郑雨军. 量子相空间纠缠轨线力学. 物理学报, 2013, 62(21): 213401. doi: 10.7498/aps.62.213401
    [12] 陈卫东, 刘要龙, 朱奇光, 陈颖. 基于改进雁群PSO算法的模糊自适应扩展卡尔曼滤波的SLAM算法. 物理学报, 2013, 62(17): 170506. doi: 10.7498/aps.62.170506
    [13] 欧阳成, 姚静荪, 温朝晖, 莫嘉琪. 一类广义鸭轨迹系统轨线的构造. 物理学报, 2012, 61(3): 030202. doi: 10.7498/aps.61.030202
    [14] 盛峥. 扩展卡尔曼滤波和不敏卡尔曼滤波在实时雷达回波反演大气波导中的应用. 物理学报, 2011, 60(11): 119301. doi: 10.7498/aps.60.119301
    [15] 李荣, 伍歆. 两个三阶最优化力梯度辛积分器的对称组合. 物理学报, 2010, 59(10): 7135-7143. doi: 10.7498/aps.59.7135
    [16] 郑世旺, 贾利群. Birkhoff系统的局部能量积分. 物理学报, 2006, 55(11): 5590-5593. doi: 10.7498/aps.55.5590
    [17] 刘思敏, 张光寅, 荣放, 吴斯嘉, 康玮, 刘秋香. 固液同成分点组分的LiNbO3晶体吸收边的异常紫移. 物理学报, 1985, 34(2): 275-279. doi: 10.7498/aps.34.275
    [18] 刘辽. 费曼路径积分和霍金蒸发. 物理学报, 1982, 31(4): 519-524. doi: 10.7498/aps.31.519
    [19] 王仁川. Airy函数乘积的积分. 物理学报, 1981, 30(1): 74-83. doi: 10.7498/aps.30.74
    [20] 鲍诚光. N体散射——(Ⅰ)跃迁振幅的代数结构和积分方程组. 物理学报, 1975, 24(3): 215-222. doi: 10.7498/aps.24.215
计量
  • 文章访问数:  4215
  • PDF下载量:  182
  • 被引次数: 0
出版历程
  • 收稿日期:  2017-02-27
  • 修回日期:  2017-05-01
  • 刊出日期:  2017-08-05

/

返回文章
返回