油气输运管道缺陷漏磁检测量化技术研究

通过实测得到油气输运管道缺陷漏磁(MFL)信号，分析了缺陷几何尺寸与信号特征量之间关系。采用特征提取和模式识别技术对缺陷进行量化分析。对缺陷的长、宽、深三个指标分别应用不同的特征量和相应的非线性方法进行定量识别。试验结果表明，缺陷长度、宽度和深度的预测准确度分别达到了100％、89％和77．8％。这里通过使用长宽比特征量描述方法，有效地提高了深度的估计精度，很好地解决了管道缺陷的量化识别问题。

The Quantitative Recognition for Pipe Pits on Oil-gas Pipe Magnetic Flux Leakage Inspection

The signals of pipe magnetic flux leakage (MFL) are collected by experiment. The relation between the measure of pipe pits and the signal features is studied. Feature and model recognition technology are used in quantitatively recognizing to pipe pits. The different features and nonlinear methods are applied for the analysis to the length, width and depth of pipe pits. The experiment results prove that the evaluation accuracy to the length, width and depth of pits reaches to 100%, 89%and 77.8% respectively. By means of the feature for the ratio of length-to-width,the depth evaluation accuracy is improved. This method is quite effective for quantitative recognition of pipe pits.