Extraction method of typical features of natural gas pipeline corrosion defects

Abstract

The invention discloses a method for extracting typical features of corrosion defects of natural gas pipelines, belonging to the technical field of oil and gas pipelines. It includes the following steps: collecting design parameters and conveying medium information of pipelines with defects; detecting pipeline defects by using a magnetic flux leakage detector; pre-classifying the corrosive defect data; Curve fitting to obtain the optimal distribution type and probability density function that the defect size variable obeys; a clustering algorithm based on statistical outlier detection method and roulette algorithm idea improvement is used to cluster and analyze the defect data, and obtain Defect typical size, circumferential distribution orientation characteristic value. The invention improves the selection method of the initial cluster center, reduces the adverse influence of outliers on the cluster analysis, and provides a set of scientific and comprehensive methods for extracting corrosion defect distribution law, typical size and circumferential distribution orientation characteristics from a large number of detection data. systematic approach.

Description

technical field [0001] The invention belongs to the technical field of oil and gas pipelines, in particular to a method for extracting typical features of corrosion defects in natural gas pipelines. Background technique [0002] In the oil and gas pipeline industry, corrosion is one of the main factors that threaten the safety of natural gas pipelines. It can lead to the formation of a large number of corrosion defects such as thinning and cracks on the surface of the pipeline, burying hidden dangers for the safe operation of oil and gas pipelines. Regularly inspecting and evaluating the metal surface of oil and gas pipelines, and formulating reasonable repair measures, has become an important technical means for the oil and gas pipeline industry to deal with the threat of pipeline corrosion and reduce the risk of corrosion failure. [0003] Defect detection of metal pipelines usually produces a large amount of detection data, including information such as types and sizes of...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the problems that the clustering analysis is easily disturbed by outliers and initial cluster centers, and the existing detection data processing methods cannot efficiently and reasonably extract the typical characteristics of pipeline corrosion defects. Brand and natural gas composition information, based on the statistical outlier detection method and the improved clustering algorithm based on the roulette algorithm, the method of extracting typical characteristics of corrosion defects from a large number of detection data

Images