Composite Detection Method for Subsurface Defects in Gas Gathering Station Pipelines Using Eddy Current and Ultrasonic Techniques

The gas gathering station is the first hub of the entire natural gas pipeline transmission. Some of the pipelines at the gas gathering station have wall thickness thinning, perforation and other damages, which affect the safe production operation of the gas gathering station. For a single pulse eddy current testing (ECT) or ultrasonic testing (UT) classification accuracy and detection range is limited and other problems, this paper investigates the ECT and UT composite detection method. In this paper, an optimized Dempster-Shafter(D-S) evidence fusion method based on fuzzy C-mean (FCM) clustering is proposed. First, the feature parameters of the ECT and UT response signals are extracted separately and the affiliation matrix is obtained by the FCM clustering method as the respective Basic probability assignment (BP A) of detection; Then, a solution to the problem of conflicting evidence in the composite process of heterogenous sensors is proposed to amend the ECT data by using the clustering centers obtained from the UT feature. Finally, effective classification of defects with different depths is achieved by fusing the BP As of ECT and UT using D-S evidence. The results show that compared with the traditional D-S evidence fusion method based on Bayes transform and single-sensor detection technique, the detection method proposed in this paper has improved the identification accuracy and detection range of subsurface defects.