Application of Neural Network to Diesel Engine SOA

Author

Hongxiang, Tian ; Yuntao, Liu ; Xiangjun, Wu

Author_Institution

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

Volume

1

fYear

2011

fDate

6-7 Jan. 2011

Firstpage

555

Lastpage

558

Abstract

In order to mine further the atomic emission spectroscopic data of diesel engine oil, two groups of oil were taken. The first group contained 33 oil samples made up CD40, CF40 diesel engine oil and hydraulic oil with wear metal debris and contaminants. The second group contained 60 oil samples separately from running condition and routine condition. With the rotary disk electrode (RDE) spectroscopic instrument, 21 elements´ concentrations of every oil sample were detected. To identify the kinds of oil and diesel engine working conditions, neural network analysis methods were applied to all spectroscopic data. The results shows that mining oil spectroscopy data by neural network analysis methods can be used to reveal the kinds of oil and to distinguish diesel engine routine condition from running condition.

Keywords

automobile industry; contamination; diesel engines; hydraulic fluids; neural nets; spectroscopy; wear; CD40 diesel engine oil; CF40 diesel engine oil; RDE spectroscopic instrument; atomic emission spectroscopic data; contaminants; diesel engine routine; hydraulic oil; neural network analysis; oil samples; oil spectroscopy data; rotary disk electrode; routine condition; running condition; wear metal debris; Additives; Artificial neural networks; Diesel engines; Metals; Petroleum; Predictive models; Spectroscopy; Diesel Engine; Lubricating Oil; Neural Networks; Spectrometric Oil Analysis (SOA);

fLanguage

English

Publisher

ieee

Conference_Titel

Measuring Technology and Mechatronics Automation (ICMTMA), 2011 Third International Conference on

Conference_Location

Shangshai

Print_ISBN

978-1-4244-9010-3

Type

conf

DOI

10.1109/ICMTMA.2011.141

Filename

5720844