The Study Of Signal Restruction Of Magnetic Flux Leakage Inner Detection For Pipeline

Oil and gas are the main energy in our country. The way of transporting energy is generally in pipeline. The environment of pipeline is complex because pipeline is berried under ground or undersea for years. So, different level of defects such as corrosion, distortion and other damage and human destroy. Also there is human destruction. It may cause leak that not only pollutes environment but also make country economic loss. So pipeline detection plays an important role in it. MFL inner detection in pipeline is an advanced detection technology home and abroad. MFL inner detection device can be used to detect the pipeline and check out defect signal effectively. For the reconstruction problem of MFL detection signal, this paper makes an intensive study.For the reconstruction problem of MFL signal, the study had been done in deep. The paper makes 2D finite element simulation of MFL detection by software named ANSYS. According to actual situations, more realistic semi-ellipse defect model is established on the pipeline wall which is more easily corroded by outer factors. By the analysis on radial magnetic flux density curve of detect, peak to peak value of magnetic density and horizontal distance difference are found. 2D simulation analysis of MFL inner detection had been operated on 90 groups of different parameter defects. 90 groups of defect data corresponding to radial magnetic flux density feature are found. Least square curve fitting method had been used to make curve fitting on simulated data, linear equation of two unknowns and binary quadric fitting equations of radial magnetic flux density feature are found. By the analysis of fitting error, binary quadric fitting equation can supply reliable math model for defect signal reconstruction.Defect signal reconstruction of math model had been done by conjugate gradient. The binary quadric fitting equations of radial magnetic flux density feature are solved by MATLAB and defect parameters of reconstructed defect are found. In contrast to solved results of Newton, conjugate gradient reconstructs defect parameters more accurately and the reconstruction rate can be above 90%. By the method of finite element simulation to get feature value and curve fitting to establish the functional relation between defect parameter and MFL feature and programming algorithm to solve nonlinear equations and implement to MFL inner detection defect signal reconstruction, satisfying results are found.