Title :
The Application Research of Pattern Recognition for Failure Detection
Author :
Li, Shufen ; Liu, Junli
Author_Institution :
Inst. of Autom., Beijing Union Univ., Beijing, China
Abstract :
Considering the characteristics of the mechanical equipment and the disadvantages of the existing diagnosis method, a new failure detection method based on pattern recognition is defined. Short time signal and linear distinguish function is adopted. Alone feature is combined for deciding. The method of sensitive and efficient is provided for detecting bearing fault. From the analysis result of the actual vibration signals, the proposed method is proven to be efficient in detecting and diagnosing the rolling element bearing faults. The method is put on application on bearing diagnosis and frication diagnosis . The results indicate that it has abilities of early stage forecast fault.
Keywords :
condition monitoring; failure analysis; fault diagnosis; flaw detection; pattern recognition; rolling bearings; signal detection; failure detection method; fault diagnosis; fault monitoring; linear distinguish function; mechanical equipment; pattern recognition; rolling element bearing fault detection; vibration signal analysis; Chemical technology; Fault detection; Fault diagnosis; Frequency domain analysis; Frequency estimation; Machinery; Pattern recognition; Rolling bearings; Signal analysis; Vibrations; Fault Diagnosis; Forecasting; Pattern Recognition;
Conference_Titel :
Services Science, Management and Engineering, 2009. SSME '09. IITA International Conference on
Conference_Location :
Zhangjiajie
Print_ISBN :
978-0-7695-3729-0
DOI :
10.1109/SSME.2009.152
