Title :
A MWPCA-based fault detection in federated no reset mode filter
Author :
Kazerooni, M. ; Khayatian, A.
Author_Institution :
Dept. of Power & Control Eng., Shiraz Univ., Shiraz, Iran
Abstract :
In this paper, two different sensor fault detection methods for federated filter structure are proposed. In the first method, measurement residual of local filters in federated no reset mode filter configuration is monitored. In the second technique, a moving window principal component analysis (MWPCA) is investigated on measurement residual of local filters outputs. Faults which are regarded in this paper include soft, hard and delta. To verify the efficiency of the proposed fault detection methods in federated filter structures, an integrated vehicular navigation system is used as a case study. Simulation results show the effectiveness and accuracy of MWPCA method with respect to measurement residual in fault detection and decreasing false alarms in federated filter structures.
Keywords :
filtering theory; principal component analysis; MWPCA-based fault detection; federated filter structure; integrated vehicular navigation system; local filters measurement residual; moving window principal component analysis; no reset mode filter; sensor fault detection method; Noise; MWPCA; federated filter; integrated navigation system; measurement residual;
Conference_Titel :
Electrical Engineering (ICEE), 2012 20th Iranian Conference on
Conference_Location :
Tehran
Print_ISBN :
978-1-4673-1149-6
DOI :
10.1109/IranianCEE.2012.6292472