Influence Of Different Hardware Parameters On Reconstruction Image In Electromagnetic Chromatography Research

Electromagnetic tomography technology is a PT technology based on the principle of electromagnetic induction.It has the strong points of non-intrusive,non-contact,and less harmful.The EMT technology can be divided into two parts:the positive problem and the inverse problem.The positive problem research focuses on obtaining the detection value and the sensitivity matrix,and the inverse problem mainly solves the image reconstruction problem.When the inverse problem is used to reconstruct the image,the minute interference in the measurement is reflected by the sensitivity matrix.And the sensitivity matrix is an important index that affects image quality,and it is an important basis for inverse problem image reconstruction.In this paper,the research significance,current situation and application value of EMT technology are systematically analyzed,the working principle of the system is introduced,and the research methods of positive and inverse problems are expounded.According to the development of the technology at home and abroad,the corresponding mathematical model is studied.The physical model of the EMT technology is determined.Based on the traditional "O" sensor array structure,a simulation model is established.The corresponding physical model is constructed through COMSOL finite element simulation software,the object field is divided into multiple tiny cells,the second-order partial differential equation is solved through the finite element method,the detection value is obtained,and the sensitivity model is established.In the sensitivity solution,the simulation perturbation method is used to explore the influence of the number of coils in the EMT sensor on the sensitivity.Using the separation variable method,compare the sensitivity matrix of sensor arrays with 8coils,12 coils,16 coils,and 20 coils.Use MATLAB software to compare the sensitivity matrices of different coils.It is concluded that in different numbers of coil sensor models,the sensitivity bands are all distributed along the arc between the excitation and detection coils.The sensitivity bands of the coils at the same position have the same shape,showing a tendency of being strong in the middle and gradually weakening to the outside.The sensitivity bands of adjacent positions gradually decrease and tend to remain unchanged.With the increase of the number of coils,the same sensitivity band widens,but the sensitivity value of the central part is slightly weakened,and the sensitivity band of adjacent positions gradually decreases and tends to remain unchanged.Determine the most suitable array structure,give a normalized model,process the sensitivity matrix of the most suitable coil sensor number array structure,and obtain the final sensitivity distribution.The normalized sensitivity distribution is more uniform than before.According to this characteristic,a certain foundation is laid for the practical application of EMT technology and image reconstruction.The system studies the basic principles of four image reconstruction algorithms,namely LBP algorithm,regularization algorithm,iterative algorithm and conjugate gradient algorithm and corresponding image evaluation indicators.Using the controlled variable method,explore the influence of the hardware parameters of the object field,including the excitation current,frequency,coil diameter,and number of turns on the image reconstruction effect.The image is reconstructed by four reconstruction algorithms,and the comparative analysis draws the conclusion:the larger the excitation current,the higher the excitation frequency,the more the sensor coil turns,the smaller the diameter,the smaller the image error index,and the higher the correlation.The sensor parameters must fully consider the impact of load and electromagnetic interference,and select appropriate parameters.This article gives the approximate range for selecting sensor parameters,which provides a basis for the design of the hardware system.