| With the development of society,the demand of railway transportation is constantly increasing,high-speed railway is the main direction of the development of railway transportation,and high-speed railway is gradually favored by people.Rail is the basis of high-speed railway rolling stock operation,and its status is directly related to the safety of railway transportation.Therefore,it is of great significance to study the non-destructive testing technology of rail for its safety detection,so as to ensure the safe operation of the train.The nondestructive testing of rail usually uses to identify the physical information of the detection target,then locate the defect location,and then through imaging to achieve the detection of specific damage on the surface and shallow surface.Among them,it is common to use the reflection and scattering information of electromagnetic wave to realize the detection process,but the technology is not mature in the sensitivity,high speed,and real-time imaging of the results.That is,the limitation of numerical calculation to achieve efficient real-time monitoring of rail is obvious.At present,many numerical methods can be used to accurately calculate the electromagnetic characteristics of targets in complex environment.It is difficult to achieve accurate and efficient analysis of practical problems,mainly due to the computational complexity and data storage.(1)Due to the influence of environment,climate,soil and other factors,the complex electromagnetic environment of high-speed railway subgrade increases the computational complexity of simulation;(2)in the process of rail scanning,it is necessary to detect the target in different frequency and direction,which results in a large amount of simulation calculation.The huge amount of calculation will cause the lack of simulation research data,which hinders the development of high-speed rail detection technology.Therefore,it is necessary to study the related efficient algorithms.The purpose of this paper is to design accurate and efficient calculation methods,simulate and analyze the parameters and indexes of high-speed rail flaw detection,and provide theoretical basis,simulation methods and analysis data for high-speed rail flaw detection.In order to obtain the electromagnetic characteristics of high-speed rail more accurately and efficiently,this paper uses the surface integral equation method based on the Moment method combined with a variety of fast algorithms,to study and analyze the rail scattering problem,so as to achieve the accurate solution and fast calculation of the problem.First of all,this paper introduces the integral equation method for solving electromagnetic scattering problems and the Moment method for solving the integral equation.On the original platform of Moment method,how to use MLFMA to accelerate the solution ofelectromagnetic characteristics of the target is introduced.For the realization of MLFMA,the matrix vector multiplication is considered as the interaction of equivalent current or magnetic current when the scattering problem is calculated by the method of Moment.Among them,Mo M and MLFMA take on different parts in the solution process,involving different processing methods of far-field and near-field.Then,an analysis method is proposed to solve the scattering problem of compound target in several different cases.Taking rail as an example,from the analysis of the self action and interaction of the target,it belongs to the multi-target problem,and there is no mutual inclusion relationship.The electromagnetic mutual coupling relationship of the rail is established.The impedance matrix Z is composed of the self action of the target and the mutual coupling between the targets.The solution of scattering problem among different target material properties is explained,including the analysis of perfectly electrical conductor,pure medium and mixed target.Finally,a hybrid model based on ACA is built.For the near-field region of the target self acting,MLFMA accelerates the matrix vector multiplication operation in the Moment method,decreases the times of matrix vector multiplication,and the storage and complexity of the calculation;and for the far-field region,it is precisely because the impedance matrix produced by the interaction between the targets has low rank characteristics,and ACA is used to compress and accelerate the filling of the matrix. |