Research On Key Issues Of High-speed Electromagnetic Tomography And Its Application

Electromagnetic Tomography(EMT)has been developed since later 1990 s,which is one of process tomography technologies based on the principle of electromagnetic induction.The technique of EMT acquires wide attentions in biomedical field,non-destructive evaluation and multi-phase flow detection et al,because of the advantage of non-invasive,non-contact,non-radiation,low-cost and real-time et al.With the increase of production efficiency in various application area,demand of real-time characteristic of EMT system is put forward by industry field.Railway transportation has the advantages of clean,energy saving,high efficiency,environment friendly and large traffic volume et al.Thus,railway transportation plays an important role in the rapid development of national economy and modernization drive.Rail track is an important part of rail system and the monitoring of rail service status has great implication to the safety of railway transportation.With the improvement of operating mileage,operating speed and operating efficiency of railway networks in recent years,there are higher standards are demanded in high-speed,on-line,accurate and visual detection of rail inspection.This thesis focuses on the demodulation algorithm and image reconstruction algorithm to improve the efficiency of EMT system.The feasibility and effectiveness of using electromagnetic tomography demodulation technique and image reconstruction technique in rail inspection are also researched.The main works has been finished as following:(1).A dual filter structure demodulation algorithm based on FPGA is proposed.The mathematical deduction of the algorithm is discussed in detail.The demodulation performance of the algorithm is a studed through theoretical analysis and the performance of the algorithm is proved through simulation by using MATLAB.A demodulation experiment system based on FPGA is built to verify the proposed demodulation algorithm.Experiment results,simulation results and theoretical analysis shows that the proposed algorithm has a certain anti-noise performance and the proposed algorithm can improve demodulation efficiency of EMT system which using data reuse technique.(2).The effectiveness of rail inspection method based on electromagnetic detection method with different detection speed are verified through experiment.An implementation method of defect inspection system that based on principle of electromagnetic induction,demodulation algorithm of EMT and Field Programmable Array(FPGA)is proposed to satisfy the requirements of real-time and rapidity of high-speed rail defect inspection.The experimental results of rail defect detection shows that the method has good detection effect under speed condition of manual,20 km/h,60km/h,120 km/h,180 km/h and 240 km/h.(3).A fast image reconstruction algorithm based on dimension reduction of sensitivity matrix is proposed.For this algorithm,firstly,matrix transformation is used to decrease the dimension of sensitivity matrix.Then,traditional image reconstruction algorithm is used to reconstruct distribution of medium.The simulation and experiment results shows that,the calculation load of reconstructed image can be decreased with limited descent of quality of reconstructed image when using proposed algorithm.(4).An optimized Landweber iteration fast image reconstruction algorithm is proposed.For this algorithm,firstly,the dimension of sensitivity matrix is reduced through matrix transformation to reduce the calculation load of each iteration.Secondly,the dimension reduced sensitivity matrix is optimized to improve the quality of reconstructed image.Finally,image is reconstructed by using Landweber iteration algorithm and pre-processed sensitivity matrix.Experiment of simulation results shows that,the proposed algorithm increase the quality of reconstructed image of Landweber iteration algorithm and decrease the calculation load of image reconstruction of Landweber iteration algorithm.(5).An inspection method based on plane electromagnetic tomography are given.According to the structure of 75 kg/m heavy load railway and the electromagnetic character of coil,a planar sensor array was designed.The uniformity of sensitivity matrix that calculated by using different methods(field extraction method and perturbation method)are discussed and the optimal calculation method are selected.Optimize the sensitivity matrix to improve the quality of reconstructed image.Simulation and experiment results shows that,the location and shape of defects can be reconstructed effectively and detection ability of cracks and triangular pits by using proposed method are well.