지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
논문 기본 정보
- 자료유형
- 학술대회자료
- 저자정보
- 발행연도
- 2011.10
- 수록면
- 492 - 497 (6page)
이용수
초록· 키워드
오류제보하기
In this study, the aim is to detect and isolate fault in complex plants by using hierarchy structure, independent component analysis (ICA) and data fusion methods. The hierarchy strategy is used to divide a system to some agents in order to reduce the complexity of the systems as well as increase the accuracy of fault detection and isolation process. The ICA is a statistical method which is used for data reduction and feature extraction from original features. The ICA statistics I², I<SUB>e</SUB>² and SPE are proposed as on-line fault detecting strategy. Contribution plots of these statistics are used for fault identification to trigger a fault diagnosis blocks in each subsystem. A data fusion method is used to combine features in intelligent ways so as to obtain the best possible diagnostic information. In this level adaptive neuro fuzzy inference system (ANFIS) is applied to fuse the extracted data. The basic idea of proposed structure is to trigger relevant fault diagnosis blocks based on the ICA statistics fault detection. The proposed structure also provides fault prediction capability. The predictability outcome was provided because fault was detected by agents with different delays. The simulation results clearly show the advantages of proposed structure to detect isolate and predict of a fault.
목차
Abstract
Ⅰ. INTRODUCTION
Ⅱ. ICA MONITORING
Ⅲ. FUSION ARCHITECTURE
Ⅳ. FDI STRUCTURE
Ⅴ. SIMULATION
Ⅵ. CONCLUSION
REFERENCES
참고문헌 (0)
참고문헌 신청이 논문의 저자 정보
최근 본 자료
댓글(0)
0
UCI(KEPA) : I410-ECN-0101-2014-569-000912971