Study on the application of robust statistical techniques in interlaboratory comparisons for oil emission spectrometers

Author

Pei, Wen ; Tian, Hong-Xiang ; Liu, Yu

Author_Institution

Coll. of Naval Archit. & Power, Naval Univ. of Eng., Wuhan, China

Volume

3

fYear

2010

fDate

9-10 Oct. 2010

Firstpage

19

Lastpage

23

Abstract

In order to insure the reliability and the validity of the detection capability and detection data, 10 oil emission spectrometry laboratories which distribute in different areas tested 3 lubricating oil samples composed by different elements content, and interlaboratory comparisons for the detection capability of oil emission spectrometers were carried out. Based on robust statistical techniques, the detection data was evaluated and the evaluation results of robust Z-score were compared with the results of En. The results shows that the robust Z-score method had 86.67% hit rate, 7.33% failed rate, 6.00% omitted rate; while the En method had 31.33% hit rate, 66.67% failed rate, 2.00% omitted rate. the robust Z-score method is more suitable to be applied to interlaboratory comparisons for oil emission spectrometers than En method.

Keywords

laboratory techniques; lubricating oils; spectrometers; statistical analysis; Z-score method; lubricating oil samples; oil emission spectrometers; robust statistical techniques; Copper; Iron; Lead; Nickel; Silicon; Tin; Zinc; En; interlaboratory comparisons; robust Z-score; spectrometer;

fLanguage

English

Publisher

ieee

Conference_Titel

Future Information Technology and Management Engineering (FITME), 2010 International Conference on

Conference_Location

Changzhou

Print_ISBN

978-1-4244-9087-5

Type

conf

DOI

10.1109/FITME.2010.5654766

Filename

5654766