Application of L-D factorization in calculating inverse in federated filter
-
摘要: 求逆是联邦卡尔曼滤波算法中数据融合的关键步骤.它的计算精度的高低和运算速度的快慢将直接影响到整个滤波算法的成功与否.经过数字仿真,把矩阵分析中的L-D分解算法运用到联邦卡尔曼滤波中的矩阵求逆当中能得到良好的效果,不但运算速度和精度有提高,整个滤波算法的容错性也有明显改善.Abstract: How to correctly calculate the inverse is one of the most important processes in federated filter. The precision of its result and time it will consume would decide whether the whole filter can work properly. The result of digital simulation illustrate that applying L-D factorization which is famous in matrix theory into federated filter to get the correct result of inverse can not only improve the speed and precision of algorithm but meliorate its error toleration.
-
Key words:
- inverse /
- digital simulation /
- Kalman filtering /
- federated filter /
- L-D factorization
-
[1] 赵亦林(美).车辆定位导航系统[M] 北京:电子工业出版社,1999 Zhao Yilin(American). Vehicles' position and guidance system[M] Beijing:Publishing House of Electronic Industry, 1999(in Chinese) [2] 吴小强,李 鹏,曲卫民(翻译).智能交通系统研究回顾与展望[J] 国外公路,2000,20(4):36~39 Wu Xiaoqiang, Li Peng, Qu Weimin(translator). Retrospect and prospect in ITS research[J] Journal Foreign Highway,2000,20(4):36~39 [3] 秦永元,张洪钺,汪叔华.卡尔曼滤波与组合导航原理[M] 西安:西北工业大学出版社,1998 Qin Yongyuan, Zhang Hongyue, Wang Shuhua. Kalman filter and combination guidance principle[M] Xi'an:Northwestern Polytechnical University Press,1998(in Chinese) [4] Stephen C A. Covariance analysis technique for the federated Kalman filter[J] IEEE,1993 [5] Lawrenece P J. Berarducci M. Comparison of federated and centralized Kalman filters with fault detection considerations[J] IEEE, 1996 [6] 胡涛松,邹志情.大系统的分散滤波结构及容错应用[J] 数据采集与处理,1994,9(3):1~8 Hu Taosong, Zou Zhiqing. Application of distribution filter structure and error toleration technology in big system[J] Journal Data Acquisition and Processing,1994,9(3):1~8(in Chinese) [7] 高钟毓.工程系统中的随机过程[M] 北京:清华大学出版社,1989 Gao Zhongyu. Stochastic process in engineering system[M] Beijing:Tsinghua University Press,1989(in Chinese) [8] Broatch S A, Henley A J. An integrated navigation system manger using federal Kalman filtering[J] IEEE, 1991 [9] Lawrence P J, Berarducci M P. Navigation sensor, filter and failure mode simulation results using the distributed Kalman filter simulator(DKFSIM)[J] IEEE, 1994 [10] 叶中付,沈凤麟,王 涛.基于L-D分解的广义特征值DOA估计方法[J] 重庆邮电学院学报,1995,7(1):23~31 Ye Zhongfu, Shen Fenglin, Wang Tao. DOA estimation based on generalizeg eigenvalues of L-D factorization of correlation matrix[J] Publishing House Journal of Chongqing University of Posts and Telecommunications,1995,7(1):23~31(in Chinese)
点击查看大图
计量
- 文章访问数: 3051
- HTML全文浏览量: 247
- PDF下载量: 1198
- 被引次数: 0