Study On Nonmagnetic Pipeline Magnetic Flux Leakage Detection Technology Based On AC Excitation Method

The traditional non-destructive testing methods mainly include the permanent magnet flux leakage testing method and the DC flux leakage testing method.Because of the stable magnetic field of both detection methods,it is convenient for us to analyze and study them,but they all have disadvantages.The permanent magnet excitation flux leakage detection model of the excitation source is fixed,so its size can not be changed,but also because of permanent magnet excitation,the detector model is usually larger volume.The magnetic field leakage detection method of DC excitation method can be changed accordingly.However,DC magnetization usually requires a large current when magnetized the tested object,so it has the disadvantage of excessive power consumption.Aiming at the defects of the permanent magnetic flux leakage non-destructive testing method and the DC leakage magnetic testing method,this thesis presents a new AC leakage magnetic non-destructive testing method.It has a series of advantages such as small size and low power consumption,and has a wide range of applications.This article mainly studies the following questions.Aiming at the problem of complex and difficult to analyze alternating magnetic field,a three-dimensional finite element simulation model based on ANSYS software was established to analyze the characteristics of alternating magnetic field distribution and eddy current field.At the same time,the value of AC,Geometric parameters and other factors on the magnetic flux leakage of the law,in order to achieve high-precision detection of finite element analysis.Aiming at the problem of the influence of frequency on the MFL signal with different size defects,this thesis presents a method to choose the best detection frequency by analyzing the skin effect characteristic,quantifying the depth of the skin layer,finding the effect of frequency on MFL signal,and obtaining the different size defects.Aiming at the problem that the feature value of the depth value of the defect can not be evaluated from the finite element simulation analysis,a hybrid method of defect contour reconstruction based on differential evolution and tabu search is proposed.Using radial basis function neural network as the forward model,a large number of contour-signal samples are used to train the forward model of RBFNN,and the differential evolution and tabu search hybrid algorithm are used to update the estimated defect contour reconstruction.The simulation results show that this method is used to reconstruct Defective contour,both in the contour accuracy and reconstruction contour speed has significantly improved.