Research On Fast Calculation Of Magnetic Flux Leakage Signal And Defect Inversion Method Of Ferromagnetic Materials

Ferromagnetic materials are commonly used in the manufacture of electrical equipment,such as transformer core,generator motor rotor,steel core aluminum strand,etc.After a long time of using ferromagnetic materials,there will be various defects,which will affect the normal use of equipment.The defect detection of ferromagnetic materials has practical significance,and defects can be found as early as possible to avoid unnecessary losses.The defective ferromagnetic materials will produce leakage magnetic field under the action of external magnetic field,so the information of defects can be determined by magnetic leakage detection.The main purpose of magnetic leakage detection is to calculate the shape and size of defects by the measured magnetic leakage signal.In view of the complex calculation of the existing magnetic leakage signal,a fast calculation method of magnetic flux leakage signal based on the reduction of the numerical calculation area based on the magnetic disturbance characteristics is proposed,and the inversion calculation of the defect shape is carried out based on the method.Finally,the calculation method of defect size is studied for the practical needs of engineering.Firstly,a two-dimensional model of magnetic leakage detection is established and numerical calculation is carried out in COMSOL.The change law of magnetic leakage signal is studied when the defect size is changed;the characteristics of x-axis and y-axis components of different defects magnetic leakage signals are analyzed,and the characteristic quantity that can reflect the magnetic leakage signal is determined,which provides data support for the research of defect size.Secondly,the distribution characteristics of magnetization intensity of ferromagnetic materials after magnetization are analyzed.The distribution chart of x-axis component of magnetization intensity which can reflect the information of defect shape is obtained,which provides theoretical basis for inversion of defect shape.In order to solve the problems of the integral forward modeling of magnetic dipole elements,a method of reducing the calculation area based on the magnetic disturbance characteristics is proposed.The original forward model is improved.The results of the calculation examples show that the method improves the accuracy of the calculation results and reduces the calculation timeThirdly,the inversion research is carried out from the two aspects of defect shape and size.For the inversion of defect shape,from the point of inverse of forward model,a new method of defect shape inversion based on improved truncated singular value decomposition method is proposed.For the inversion of defect size,a RBF neural network model based on gradient descent method is proposed.The model takes the characteristic quantity of magnetic flux leakage signal as input and the defect size as output,and realizes the direct calculation of defect size.Finally,taking the transformer core defect as the research object,the magnetic leakage detection experimental platform is built,and the inversion algorithm of defect shape and size is verified by experiments.The experimental results verify the effectiveness and accuracy of the above two inversion algorithms.