基于支持向量机算法的注水管道剩余寿命预测

支持向量机是在统计学习理论的基础上发展而来的一种新的模式识别方法, 在解决有限样本、非线性及高维模式识别问题中表现出许多特有的优势。鉴于此, 针对注水管道的使用寿命和腐蚀影响因素之间复杂的映射关系, 在注水管道的剩余寿命预测研究中引入基于统计学习理论的支持向量机算法。研究了胜利油田某实验区注水水质腐蚀的影响因素, 应用LibSVM软件建立了注水管道的剩余寿命预测模型, 从而提供了一种预测注水管道剩余寿命的新方法。实际应用结果表明, 用支持向量机算法预测注水管道剩余寿命在样本有限的情况下具有明显优势。

Predicting residual life of water injection pipeline based on support vector machines

Support vector machine (SVM) is a new pattern recognition method developed based on statistical learning theory, which has many advantages in solving the problems with finite sample, nonlinear and high dimensional pattern recognition. According to the complicated reflection relation between the service life and corrosion affecting factor of the water injection pipeline, the factors influencing the quality of the injected water in a test zone in Shengli Oilfield are investigated by means of the SVM algorithm. The model for predicting the residual life of the water injection pipeline is set up by using the LibSVM software. Practice proves this new life prediction method has obvious advantage under the condition of finite samples.