A Fault Diagnosis of Suck Rod Pumping System Based on SVM

Abstract:

Aiming at the defects of fault diagnosis in the traditional method for sucker rod pump system, a new method based on support vector machine (SVM) pump fault diagnosis is proposed. Through studying the theory of invariant moment and the shape characteristics of pump indicator diagram, seven invariant moments is extracted from the indicator diagram as a pumping unit well condition of the characteristic parameters. Then these parameters are pretreatment, and it makes up seven eigenvector which are regarded as the input eigenvector of the SVM. The experiment indicates that the method can not only detect the fault of the pumping oil well but also can recognize the fault type of it, which is very effective for safety protection and fault diagnosis of the pumping oil.

• Periodical: Applied Mechanics and Materials (Volume 307)
• Pages: 285-289
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.307.285
• Online since: February 2013
• Authors: Wei Wu, Yu Zhou, Hang Xin Wei
• Keywords: Fault Diagnosis, Feature Vector, Indicator Diagram, Invariant Moment, Support Vector Machine (SVM)

References

[1] Z.H. Cui: Rod Pumping System (Petroleum Industry Press, Chinese 1994).

[2] F.H. Shu: Power Generating Equipment, Vol. 34 (2006), p.57.

[3] Z.Q. Li, H.S. Li: Oil-Gasfield Surface Engineering, Vol. 09 (2008), p.3.

[4] X.J. Kuang, W.Q. Zhao and M.F. Li: Manufacturing Automation, Vol. 15 (2012), p.65.

[5] X.J. Liu, H.X. Pan: Design and Manufacture of Diesel Engine, Vol. 03 (2011), p.21.

[6] S.F. Ding, B.J. Qi, H.Y. Tan: Journal of University of Electronic Science and Technology of China, Vol. 40 (2011), p.2.