An extension to the Kalman filter for an improved detection of unknown behavior

被引:3
作者
Benazera, E [1 ]
Narasimhan, S [1 ]
机构
[1] NASA, Ames Res Ctr, QSS Grp Inc, Moffett Field, CA 94035 USA
来源
ACC: Proceedings of the 2005 American Control Conference, Vols 1-7 | 2005年
关键词
D O I
10.1109/ACC.2005.1470097
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The use of Kalman filter (KF) interferes with fault detection algorithms based on the residual between estimated and measured variables, since the measured values are used to update the estimates. This feedback results in the estimates being pulled closer to the measured values, influencing the residuals in the process. Here we present a fault detection scheme for systems that are being tracked by a KF. Our approach combines an open-loop prediction over an adaptive window and an information-based measure of the deviation. of the Kalman estimate from the prediction, to improve fault detection.
引用
收藏
页码:1039 / 1041
页数:3
相关论文
共 6 条
[1]  
Cover T. M., 2005, ELEM INF THEORY, DOI 10.1002/047174882X
[2]  
DEFREITAS N, 2002, IEEE AEROSPACE
[3]   Characterization of Kalman filter residuals in the presence of mismodeling [J].
Hanlon, PD ;
Maybeck, PS .
IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS, 2000, 36 (01) :114-131
[4]  
NAYAK P, 2000, P AAAI 2000
[5]  
RUDOLPH EK, 1960, T ASME, V82, P35
[6]  
2004, P 15 INT WORKSH PRIN