DocumentCode
2451723
Title
Transit Articles Extraction Based on Domestic Fusion Algorithm
Author
Wang, Mingyi ; Wang, Liqi
Author_Institution
Sch. of Meas. & Commun., Harbin Univ. of Sci. & Technol., Harbin, China
fYear
2009
fDate
25-26 April 2009
Firstpage
785
Lastpage
787
Abstract
An elaborately designed software architecture is put forward based on fuzzy sets theory (FST), which is specialized in multiple sensor fusion and mechanism failure diagnosis. Besides, when confronted with multiple fault signals, fusion parameters can be dynamically adapted based on principles of fuzzy soft clustering so as to promote immune ability in artificially mechanical systems. The key point in this new approach lies in its power on faults detection, which requires no prior information on the state vectors of the sensors and system behavior, and no supplemental machine learning operation is required. The proposed algorithm combines principles of artificial immune system and the classical technique in fuzzy theory, which will consist of two main portions. In the first part a traditional data fuse structure is constructed, the sensor signals will be fed into it to implement the fuzzy aggregating algorithm.
Keywords
fault diagnosis; feature extraction; fuzzy set theory; learning (artificial intelligence); pattern clustering; sensor fusion; software architecture; artificial immune system; artificially mechanical system; domestic fusion algorithm; faults detection; fuzzy aggregating algorithm; fuzzy sets theory; fuzzy soft clustering; machine learning operation; mechanism failure diagnosis; multiple sensor fusion; software architecture; transit articles extraction; Clustering algorithms; Fault detection; Fuzzy set theory; Fuzzy systems; Mechanical sensors; Mechanical systems; Sensor fusion; Sensor systems; Software architecture; Software design; MATLAB; artificial immune; data fusion; fuzzy; mechanism system;
fLanguage
English
Publisher
ieee
Conference_Titel
Artificial Intelligence, 2009. JCAI '09. International Joint Conference on
Conference_Location
Hainan Island
Print_ISBN
978-0-7695-3615-6
Type
conf
DOI
10.1109/JCAI.2009.190
Filename
5159120
Link To Document