Purpose
This paper aims to overcome the defect that the traditional clustering method is excessively dependent on initial clustering radius and also provide new technical measures for detecting the component content of lubricating oil based on the fuzzy neural system model.
Design/methodology/approach
According to the layers model of the fuzzy neural system model for the given sample data pair, the new clustering method can be implemented, and through the fuzzy system model, the detection method for the selected oil samples is given. By applying this method, the composition contents of 30 kinds of oil samples in lubricating oil are checked, and the actual composition contents of oil samples are compared.
Findings
Through the detection of 21 mineral elements in 30 oil samples, it can be known that the four mineral elements such as Zn, P, Ca and Mg have largest contribution rate to the lubricating oil, and they can be regarded as the main factors for classification of lubricating oil. The results show that the fuzzy system to be established based on sample data clustering has better performance in detection lubricant component content.
Originality/value
In spite of lots of methods for detecting the component of lubricating oil at the present, there is still no detection of the component of lubricating oil through clustering method based on sample data pair. The new nearest clustering method is proposed in this paper, and it can be more effectively used to detect the content of lubricating oil.
[1]
Tao Liu,et al.
[Application of PCA to diesel engine oil spectrometric analysis].
,
2010,
Guang pu xue yu guang pu fen xi = Guang pu.
[2]
Jun Hong,et al.
A comparative study on simulation and experiment of oil-air lubrication unit for high speed bearing
,
2016
.
[3]
Sen Xu,et al.
Cluster Ensemble Based on Spectral Clustering
,
2012
.
[4]
Ching-Chang Wong,et al.
A hybrid clustering and gradient descent approach for fuzzy modeling
,
1999,
IEEE Trans. Syst. Man Cybern. Part B.
[5]
Li Jiexing.
An Improved Closest Cluster Learning Algorithm
,
2000
.
[6]
Xiaoping Li,et al.
Universal approximation of polygonal fuzzy neural networks in sense of K-integral norms
,
2011,
Science China Information Sciences.
[7]
Matthias Scherge,et al.
In-situ detection of hydrogen evolution in a lubricated sliding pin on disk test under high vacuum
,
2016
.
[8]
Zhang A-bu.
Design of hierarchical fuzzy systems based on subtractive clustering and adaptive neuro-fuzzy inference systems
,
2004
.
[9]
Li-Xin Wang,et al.
A Course In Fuzzy Systems and Control
,
1996
.
[10]
Jasna Hrovatin,et al.
On-line detection of incipient trend changes in lubricant parameters
,
2015
.