|Table of Contents|

Methods for Outlier Detection of Kalman Filtering(PDF)

《地球科学与环境学报》[ISSN:1672-6561/CN:61-1423/P]

Issue:
2008年第02期
Page:
204-208
Research Field:
地球信息科学
Publishing date:

Info

Title:
Methods for Outlier Detection of Kalman Filtering
Author(s):
NIE Jian-liang1; WU Fu-mei2
1. School o f Geological Engineering and Surveying, Changan University, Xian 710054, Shaanxi, China; 2. School o f Surveying and Mapping, In formation Engineering University, Zhengzhou 450052, Henan, China)
Keywords:
measurement outliers;  dynamic model error;  outlier detection;  Kalman filtering
PACS:
P207+1
DOI:
-
Abstract:
In order to detect the outliers of measurements and dynamic models, the predicted residual vector is used to construct methods to the integral error detection of measured outliers and dynamic model error. There are three types of statistics, that is, the outlier detection based on the credible observation, the reliable measurements at present epoch, and the estimations of Kalman filtering respectively. For the discrepancies of states, the estimations of Kalman filtering and the reliable measurements are used to detect dynamic model error. These methods are compared and analyzed with the data of navigation. It is proved that the outlier detection can improve the precision when measurement outliers or dynamic model errors exist.

References:

[1] T euni ssen P J G , S al zmann M A .A Recu rsi ve S lippage Test f or Us e in St ate-S pace Fi lt ering[J] .Manu scrip ta Geodaetica, 1989 , 14 :383-390.
[2] T euni ssen P J G .Qualit y C ont rol in Navigat ion S yst em s[J] . IEEE Aeros pace and Elect ronic S yst em s Magazin e, 1990 , 5(7):35-41 .
[3] Salzmann M .Real-Time Adapt ati on f or Model E rrors in Dynami c S yst em s[J] .Bul let in Geodesi gue, 1995 , 69 :81-89 .
[4] Lu G , Lachapel le G .S tati st ical Qualit y f or Kinemat ic GPS Posit ion[J] .Manuscripta Geodaeti ca , 1992 , 17 :270-283 .
[5] 陶本藻 .卡尔曼滤波模型 误差的识 别[J] .地壳 形变与 地震 , 1999 , 19(4):15-20 .
[6] 王国富 , 朱建军 .含有粗差 观测值的 自适应 滤波[J] .测 绘通 报 , 2004(4):19-21 .
[7] 何海波 , 杨元喜 .GPS 动态测量连续周跳检验[J] .测绘学报 , 1999 , 28(3):199-204 .
[8] 王甫红 , 刘基余 .星载GPS 载波相位 测量的周跳探测 方法研 究[J] .武汉大学学报 :信息科学版 , 2004 ,29(9):772-774 .
[9] 杨元喜 .动态系统的抗差 Kalman 滤波[J] .郑州测绘 学院学 报 , 1997 , 14(2):79-84 .
[10] Koch K R, Yang Y .Robu st Kalm an Fi lt er f or Rank Defici ent Observat ion Model[J] .Jou rn al of Geodesy , 1998, 72(8): 436-441 .
[11] Yang Y , H e H , Xu G .Adapt ively Robu st Fi lt erin g for Kinemati c Geodet ic Posi ti oning[J] .Journal of Geodesy , 2001 , 75(2/ 3):109-116 .
[12] 杨元喜 .论动态自适应滤波[J] .测绘学报 , 2001 , 30(4):293- 298 .
[13] 杨元喜 .Kalman 滤波异常 检验[J] .测绘科 学与工 程 , 2005 , 25(4):1-4 .

Memo

Memo:
-
Last Update: 2008-06-20