基于SVM的融合算法在管道裂纹缺陷漏磁信号特征识别的应用研究

Study on SVM-based Fusion Algorithm Identifying Pipeline Crack Flaw MFL Signal Characteristics

在研究分析支持向量机(SVM)的基础上,探讨分析了基于支持向量机的融合算法,利用建立的算法流程,以漏磁检测信号峰峰值、波峰值和波谷值3个特征量为研究对象,根据具体实验得到的特征数据库,进行了多分类器、多特征量匹配融合分析。研究证明,该算法能有效地对管道裂纹缺陷漏磁信号进行分析和辨识,具有一定的实用价值。

Basing on the study of support vector machine（SVM）,a SVM-based fusion algorithm was ana-lyzed.Through making use of the fusion algorithm process,the peak-to-peak value,wave crest value and wave hollow value ofmagnetic flux leakage signals were taken as the object of study.The research result testifies that this fusion algorithm can benefitMFL testing signal identification and buried-pipeline crack flaw detection.