Intelligent identification method for pipeline defect on basis of RBF (Radical Basis Function) neural network

Abstract

The invention provides an intelligent identification method for a pipeline defect on the basis of an RBF (Radical Basis Function) neural network, comprising the following steps of: (1) obtaining a pipeline defect flux leakage signal and a pipeline defect outline as detection data; (2) building an RBF neural network; (3) training and testing the neural network; and (4) predicting the pipeline defect outline by the tested neural network. The pipeline defect outline comprises the length, the width and the depth of the pipeline. With the intelligent identification method, finite tests are carried out, thus a pipeline defect outline prediction model is built. A computer simulation test is carried out for scientific prediction, and the pipeline defect outline can be accurately and quickly predicted.

Description

technical field [0001] The invention relates to an intelligent identification method for oil and gas pipeline defects, in particular to an RBF neural network-based intelligent identification method for oil and gas pipeline defects that can accurately and quickly identify pipeline defects. Background technique [0002] It is a difficult point in the research of magnetic flux leakage detection to reflect the characteristic parameters and contours of defects through the measured magnetic flux leakage signals. The general solution is to analyze the relationship between the defect parameters and the magnetic flux leakage signal, and calculate the relevant information of the defect according to the characteristics of the signal. The method of empirical formula is used to quantify the geometric size of defects. This method is available when the requirements for quantification accuracy are not very high, but for high-precision defect quantification, this method is difficult to meet ...

AI Technical Summary

Problems solved by technology

Quantify the geometric size of defects by using empirical formulas. This method is available when the requirements for quantification accuracy are not very high, but for high-precision defect quantification, this method is difficult to meet the requirements.

Defect reconstruction based on the inverse equation of the magnetic dipole, this method has a great advantage in identifying defects with simple shapes, but the actual defects of oil and gas pipelines are more complicated and have various forms, and the calculation amount of this method is huge

The magnetic flux leakage signal is affected by many factors. It is difficult to find a simple correspondence between the magnetic flux leakage signal and the geometric size of the defect. It is still very difficult to establish a defect identification model based on several features.

Images