Title :
Application of Principal Component Analysis to Lubricating Oil Spectral Data
Author :
Tian, Hongxiang ; Liu, Tao
Author_Institution :
Coll. of Naval Archit. & Power, Naval Eng. Univ., Wuhan, China
Abstract :
In order to monitor wear condition of marine mechanical equipment and find the cause of abnormal wear, 336 oil samples were taken from 8 steam turbine dynamos of the same type during the period of 8 years. With the Spectroil M type Instrument, elements´ concentrations of each oil sample were analyzed. The average concentrations and their standard deviation of wear, contamination and additives elements in the samples from each steam turbine dynamo were calculated. The correlation of main elements in the oil samples was analyzed by correlation coefficient analysis. Clustering of main elements in the oil from 2 of 8 steam turbine dynamos (A and G) were investigated by principal component analysis.
Keywords :
DC generators; lubricating oils; principal component analysis; steam turbines; wear; Spectroil M type Instrument; abnormal wear; lubricating oil spectral data; marine mechanical equipment; principal component analysis; steam turbine dynamos; wear condition monitoring; Condition monitoring; Educational institutions; Electronic mail; Lubricating oils; Magnetohydrodynamic power generation; Petroleum; Pollution measurement; Power engineering and energy; Principal component analysis; Turbines; condition monitoring; correlation coefficient; lubricating oil; principal component analysis; spectral analysis;
Conference_Titel :
Information Management, Innovation Management and Industrial Engineering, 2009 International Conference on
Conference_Location :
Xi´an
Print_ISBN :
978-0-7695-3876-1
DOI :
10.1109/ICIII.2009.76
