Title :
A Fault Diagnosis Method to Hydraulic Tube Tester Production Process
Author :
Wang, Shu ; Hu, Xuefa ; He, Dakuo ; Wang, Fuli
Author_Institution :
Northeastern Univ., Shenyang
Abstract :
The hydraulic tube tester process can be characterized as a data-rich process as it can generate a large number of process measurements on many variables in short time. A data mining based multiway independent component analysis and multiway fisher discriminant analysis (MICA-FDA) method is developed for fault diagnosis for the hydraulic tube tester process, where MICA models are derived to find the underlying components from normal process data, and then MFD A models are built for fault diagnosis from the various known fault data. The proposed method is applied to a hydraulic tube tester production process. Results of simulation clearly demonstrate the effectiveness and feasibility of the proposed method.
Keywords :
data mining; fault diagnosis; independent component analysis; production engineering computing; production testing; test equipment; data mining; data-rich process; fault diagnosis; hydraulic tube tester production process; multiway fisher discriminant analysis; multiway independent component analysis; process measurements; Automatic testing; Automation; Data mining; Fault diagnosis; Helium; Independent component analysis; Laboratories; Monitoring; Petroleum; Production;
Conference_Titel :
Fuzzy Systems and Knowledge Discovery, 2007. FSKD 2007. Fourth International Conference on
Conference_Location :
Haikou
Print_ISBN :
978-0-7695-2874-8
DOI :
10.1109/FSKD.2007.25