Research On Oil Pipe Defect Detection Technology In Minor Repair Operations

The oil pipe in the oil well is in a special environment.On the one hand,the oil pipe will be affected by the medium with corrosive characteristics,chemical corrosion will occur for a long time,and defects such as perforation,groove and large area thinning will occur;On the other hand,under the influence of factors such as geology and engineering,the oil pipe will undergo physical deformation,such as partial grinding,bending and cracking.Based on the magnetic flux leakage non-destructive testing technology,this paper conducts real-time detection of oil well tubing defects during minor repair operations,using a combination of finite element simulation and field experiments.The research focuses on the distribution of leakage magnetic field in the leakage magnetic flux detection of the tubing,the denoising technology of magnetic flux leakage signal and the quantitative identification of defects,and then suggestions are made for the classification and reuse of the oil pipe.Firstly,based on the principle of leakage magnetic oil pipe defect detection,the finite element simulation model is built by multi-physics coupling software for the simulation experiment.When the contour size of the tubing defect changes,the variation law of the corresponding leakage magnetic field is obtained,and the relationship between the depth,length and width of the defect and the characteristic quantity of the leakage magnetic field is determined,which provides support for the quantitative inversion of the defect.The similarities and differences of the corresponding leakage magnetic field signals when the same parameter defects are located on the inner wall of the pipeline and the outer wall of the pipeline are compared and analyzed.Secondly,the tubing defect detection system was built in the minor repair operation,realizing the real-time detection of leakage magnetic field caused by oil pipe defects.Due to the special environment of the tubing,the detection signal is polluted by various noises.In order to effectively remove the noise,the denoising of the magnetic flux leakage signal is studied.The denoising effect of the set empirical mode decomposition method and the complementary set empirical mode decomposition method on the defect magnetic flux leakage signal is compared.The results show that the denoising effect of the complementary set empirical mode decomposition method is better than the denoising effect of the set empirical mode decomposition method.Based on the empirical modal decomposition of complementary set,the eigenmode function after decomposition is analyzed.It is found that some eigenmode functions used for reconstruction also contain noise.So the idea of mutual information is introduced,and a particle swarm optimization wavelet threshold method is proposed to re-filter the noisy eigenmode function to improve the signal-to-noise ratio of the magnetic flux leakage signal.The results are compared with the wavelet threshold method.The results show that the signal-to-noise ratio of the proposed method is better than that of the conventional wavelet threshold method,which can provide favorable conditions for the quantitative inversion of tubing defects.Finally,in order to improve the quantitative inversion accuracy of oil pipe defects,the internal and external defects of tubing and the quantitative inversion of defect parameters were carried out.Aiming at the problem of distinguishing between internal and external defects of oil pipeline,the research on the differentiation of inner and outer wall defects of multi-sensor fusion tubing based on digital signal difference method is carried out.For the problem of defect quantitative inversion,the multi-output support vector regression model is introduced according to the selected feature quantity.In order to solve the problem that the parameters in the multi-output support vector regression model are difficult to determine and overcome the problem of local optimization of particle swarm.In this paper,the genetic algorithm and particle swarm optimization algorithm are combined to optimize the parameters of the defect recognition model.The results show that it can accurately invert the defect depth of the tubing defect depth of 0.5mm and above;when the defect length is long,the length inversion effect is better.