Research On Fine Imaging Method Of High Resolution Radar

In recent years,the global security situation has become increasingly deplorable.Many regions in the world have encountered terrorist attacks on civilians.Moreover,most of these attacks occurred in densely packed places such as airports,railway stations,subway stations,and stadiums.In this scenario,the traditional security measures are not effective because of the cooperative and single target checking method.Thanks to the safe,hidden and reliable imaging detection ability,high-resolution radar has become a new trend in the security check field.In particular,the fine imaging technology of high resolution radar can obtain the fine image with rich details of the target by means of large angle and multiple perspectives,which can greatly reduce the difficulty of the later image interpretation and is more suitable for the application of security check imaging.The traditional high-resolution radar fine imaging system can not meet the requirements of rapid security check and screening in high-throughput areas with dense personnel due to its expensive cost,cooperative imaging method,and less security imaging targets.Based on these backgrounds,this dissertation has studied on the theory and key technologies of fine imaging method of high-resolution radar and has carried out the researches in terms of saving the cost of radar resources,improving the real-time performance of imaging algorithm and non-cooperative imaging mode,etc.The main work and contributions include the following aspects:1.Aiming at the application requirements of low cost,high resolution and quasireal-time fine imaging of array radar system,the high resolution fine imaging method of cross array was studied.The radar imaging mode of cross array was proposed,which was constructed by receiving and sending linear perpendicular arrays to obtain the target two-dimensional scattering information.Based on the geometric model of radar imaging system,a wavenumber domain fast imaging algorithm was proposed,which useed the approximate relation of wavenumber domain to remove the radial distance term from the distance formulation.Then the two-dimension inverse fast Fourier transform(2D-IFFT)was used to speed the algorithm.Meanwhile,a correction algorithm for wavenumber domain approximate error was proposed to improve the effective imaging range with block correction method to keep its real-time performance.2.In order to solve the problem of low efficiency of security screening imaging system,multi-target circular synthetic aperture radar(CSAR)fine imaging method was studied.Firstly,a wavenumber domain imaging algorithm based on scene segmentation was proposed.In this algorithm,the multi-target imaging scene was segmented into several single-target scenes with the multi-view one-dimensional range informations to reduce the imaging range.Meanwhile,the fast Fourier transform(FFT)of angle dimension was adopted to improve its efficiency.Aiming at fast imaging of the sub-scenes,an extended polar format algorithm(E-PFA)was proposed.By compensating the phase factor of the sub-scene,a new approximate relationship of the sub-scene field was deduced,which renewed the interpolation mapping of sub-scene wavenumber domain and improved the kernel function of the PFA algorithm.It enables the PFA algorithm to perform high resolution,fast and fine imaging in any region of the target.3.For the application of non-cooperative high resolution radar fine imaging,a study on the synthetic aperture radar(SAR)fine imaging principle of moving human target was carried out,which described the three-dimensional moving target SAR imaging response and analyzed the migration,defocus and deformation mechanism of moving objects in the result.Besides,combined with the Boulic human motion model,the changes in the imaging results of the moving human body with different system parameters were analyzed,and the design guidance of the non-cooperative fine imaging system of the moving human body was given.Furthermore,a correction method of main Doppler migration of human movement was proposed,which estimated the whole velocity of target form changes of the echo range profiles in a short time to remove the Doppler migration.4.Aiming at solving the problem of Doppler migration,defocus and deformation caused by non-rigid body movement in non-cooperative imaging results,a fine imaging algorithm of SAR moving body based on error phase compensation was proposed,which assumed that the body motion in a short period of time is a rigid motion.The movement phase of the moving body torso was estimated from the peak information of fractional Fourier transform(FrFT)of small aperture echo data to correct the phase error of the torso motion.Besides,the wavenumber domain resampling was carried out to avoid the plane wave assumption phase error.With a 2D-IFFT operation,the high resolution noncooperative fine imaging of moving body was realized.