Title :
Application of statistical learning theory to predict corrosion rate of injecting water pipeline
Author :
Zhen, Huang ; Hong, Liu ; Mujiao, Fan ; Chunbi, Xu
Author_Institution :
Southwest Oil & Gas field Co., Chengdu, China
Abstract :
Support Vector Machines (SVM) represents a new and very promising approach to pattern recognition based on small dataset. The approach is systematic and properly motivated by Statistical Learning Theory (SLT). Training involves separating the classes with a surface that maximizes the margin between them. An interesting property of this approach is that it is an approximate implementation of Structural Risk Minimization (SRM) induction principle, therefore, SVM is more generalized performance and accurate as compared to artificial neural network which embodies the Embodies Risk Minimization (ERM) principle. In this paper, according to corrosion rate complicated reflection relation with influence factors, we studied the theory and method of Support Vector Machines based the statistical learning theory and proposed a pattern recognition method based Support Vector Machine to predict corrosion rate of injecting water pipeline. The outline of the method is as follows: First, we researched the injecting water quality corrosion influence factors in given experimental zones with Gray correlation method; then we used the LibSVM software based Support Vector Machine to study the relationship of those injecting water quality corrosion influence factors, and set up the mode to predict corrosion rate of injecting water pipeline. Application and analysis of the experimental results in Shengli oilfield proved that SVM could achieve greater accuracy than the BP neural network do, which also proved that application of SVM to predict corrosion rate of injecting water pipeline, even to the other theme in petroleum engineering, is reliable, adaptable, precise and easy to operate.
Keywords :
condition monitoring; correlation methods; corrosion; learning (artificial intelligence); pattern recognition; petroleum industry; pipelines; risk analysis; statistical analysis; support vector machines; water quality; BP neural network; LibSVM software; SLT; SRM induction principle; SVM; Shengli oilfield; artificial neural network; corrosion rate prediction; embodies risk minimization principle; gray correlation method; pattern recognition method; petroleum engineering; statistical learning theory; structural risk minimization; support vector machines; water pipeline injection; water quality corrosion; Corrosion; Data models; Kernel; Pipelines; Statistical learning; Support vector machines; Training; Injecting Pipeline; Prediction; Statistical Learning Theory; corrosion rate;
Conference_Titel :
Cognitive Informatics (ICCI), 2010 9th IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-8041-8
DOI :
10.1109/COGINF.2010.5599754