Research on SVM Based Diagnosis System for Oil Tubing

Abstract:

Oil tubing is one of the most used equipment in oil extraction operations. An effective diagnosis system for it can provide multifarious benefits such as improved safety, efficient production and reduced costs for maintenance. In this paper, a support vector machine (SVM) based diagnosis system for oil tubing is studied and designed. SVM method has many advantages in solving the problem of small sample, pattern recognition of high dimensionality and nonlinear problems, which is fitable to the situation of oil tubing diagnosis. The SVM based diagnosis system for oil tubing is consisted of two parts: The hardware system part, including the detector and conditioning circuit board, and the software system part, including the SVM based analysis system. The detector is disposed on the wellhead and detects the leakage of magnetic field. The conditioning circuit board focuses on signal amplification and noise removal. The SVM based analysis system diagnoses the faults by the features of detected signal. An experiment platform is designed to certify the whole system and prove it perform well with a high diagnosis accuracy.

• Periodical: Key Engineering Materials (Volume 693)
• Pages: 1405-1411
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.693.1405
• Online since: May 2016
• Authors: Lu Yang Jing, Dong Xiang Chen, Wang Tai Yong*, Peng Wang
• Keywords: Magnetic Flux Leakage Testing, Oil Tubing, Support Vector Machine, System Implement

References

[1] Wang Taiyong, Yang Tao, Jiang Qi, The quantitative recognition for pipe pits on oil-gas pipe magnetic flux leakage inspection, Acta Metrologica Sinica. 25 (2004) 247-249.

[2] Wang Taiyong, Jiang Qi, Study on quantitative recognition technology of pipeline defect, Journal of Tianjin University. 36 (2003) 55-58.

[3] Haueisen Jens, Unger Ralf, Evaluation of inverse algorithms in the analysis of magnetic flux leakage data, IEEE Transactions on magnetic, 38 (2002) 1481-1488.

DOI: 10.1109/20.999121

[4] Yang Tao, Nondestructive testing of oil-well tubing and quantitative recognition of defects based on multi-sensor fusion, Tianjin University, Tianjin, (2004).

[5] Jing Luyang, Key technology research of quantitative recognition of wire ropes injury, Qingdao Technological University, Qingdao, (2012).

[6] Jianzhong Zhou, Han Xiao, Chaoshun Li, et al. Shaft orbit identification for rotating machinery based on statistical fuzzy vector chain code and support vector machine, Journal of Vibroengineering, 16 (2014) 713-724.

[7] Yingwei Zhang, Nan Yang, Shipeng Li, Fault isolation of nonlinear processes based on fault directions and features, IEEE Transactions on Control Systems Technology, 22 (2014) 1567-1572.

DOI: 10.1109/tcst.2013.2283925

[8] Xiaoyuan Zhang, Daoyin Qiu, Fuan Chen, Support vector machine with parameter optimization by a novel hybrid method and its application to fault diagnosis, Neurocomputing, 149 (2014) 641-651.

DOI: 10.1016/j.neucom.2014.08.010

[9] Jinbao Yao, He Xia, Nan Zhang, et al. Prediction on building vibration induced by moving train based on support vector machine and wavelet analysis, Journal of Mechanical Science and Technology, 28 (2014) 2065-(2074).

DOI: 10.1007/s12206-014-0501-z

[10] Pakin Wong, Zhixin Yang, Chi Man Vong, et al. Real-time fault diagnosis for gas turbine generator systems using extreme learning machine, Neurocomputing , 128 (2014) 249-257.

DOI: 10.1016/j.neucom.2013.03.059