Research On Radar Angular Super-resolution Imaging Method Based On Deconvolution

Up to now,the technology of radar forward-looking imaging has made some achievements.However,the Wiener filtering and monopulse sharpening technology can improve the imaging quality to a certain extent,but it does not fundamentally improve the resolution.The resolution is the main factor that determines the performance of radar detection,and the study of distance resolution has been a good result.Therefore,it is of great significance to propose an algorithm to change the super-resolution capability of the radar angle.In this dissertation,radar scanning angle super-resolution imaging method based on deconvolution is researched in order to improve the radar angular resolution.This paper summarizes the principle of radar front-view scanning imaging,and analyzes and researches the problems existing in the radar angle super-resolution algorithm.Some existing imaging algorithms have been improved.The main research contents of the full text are as follows:(1)In this thesis,starting from the relative geometric relationship between the radar platform and the observation target,the echo signal model is studied and derived to provide the basis for the follow-up super-resolution algorithm research.The performance of the deconvolution solution objective is affected by the inherent morbidity and noise sensitivity of the deconvolution.Therefore,this paper considers the introduction of a reasonable regularization constraint pair to mitigate the noise sensitivity problem,while considering the introduction of a multi-channel model.To ease the morbidity of deconvolution.In this paper,a Bayesian forward-looking super-resolution imaging method based on Shannon's entropy transcendental multi-channel missile-borne radar is proposed.This method is based on missile-borne monopulse scanning radar.It first uses the information of the sum-difference observation channel to alleviate the pathological state of the imaging solution,thus establishing a sum-difference two-channel echo model.Then,the entropy is introduced to describe the target scattering prior information.A Bayesian maximum a posteriori estimate is used to establish a combined deconvolution front-view imaging optimization solution model.Finally,the improved conjugate gradient method is used to optimize the imaging.Simulation results show that compared with the improvedmulti-channel Wiener filtering algorithm and LR algorithm,the proposed method can obtain higher accuracy,and has better anti-noise performance,which can effectively suppress the generation of false targets.(2)Considering the problem that multiple parameters need to be manually adjusted in the super-resolution algorithm based on Shannon's entropy priors,an adaptive parameter updating method based on confidence framework is proposed.From the belief frame theory,when the initial value of the target scattering information is given,the parameter updating formula obtained through the confidence framework theory does not need to artificially set the unknown parameters such as the regularization factor and the noise variance,so that during the solution process of the algorithm,Unknown parameters are automatically updated to the appropriate values,effectively improving computational efficiency.(3)In radar real-beam scanning,single-pulse sharpening techniques can improve imaging quality due to radar aperture limitations and other issues,and there are certain problems in achieving super-resolution.To solve this problem,this paper deduces a maximum posterior super-resolution algorithm based on Tsallis entropy.This method is based on single-pulse scanning radar.Tsallis entropy is introduced as a priori information to mitigate noise sensitivity and other problems.A forward-looking imaging optimization model is established by Bayesian maximum a posteriori estimation.The simulation results show that the flexibility of Tsallis entropy can effectively improve the efficiency and super-resolution of the super-resolution algorithm.(4)In the original MMSE method,no noise was effectively processed.To solve this problem,this paper proposes an improved iterative adaptive super-resolution algorithm.This method is based on the minimum mean square error model.Firstly,the original model is improved.Then the Tsallis entropy is introduced as a regularization constraint term to establish a deconvolution front-view imaging optimization model.Iterative adaptive method to solve.The simulation results show that the proposed method can improve the estimation accuracy and enhance the super-resolution effect.