Research On The Acceleration Algorithm Of Forward-Looking Beam Sharpening For Airborne Radar

Airborne scanning radar can obtain real-beam images of the front and lower areas of the flight platform.It has important application prospects in forward-looking reconnaissance,autonomous navigation,and terminal guidance.However,the azimuth resolution of real-beam images has always been limited by the antenna size and radar function,and it is usually poor.Beam sharpening technology can achieve azimuth super-resolution by using a priori information and convolutional inversion operation,but it faces the problem of slow convergence.This article focuses on the research of the airborne forward-looking scanning radar beam sharpening acceleration algorithm.Through establishing an echo model,analyzing the echo law,studying iterative threshold shrinkage based on regularization theory,and maximum posterior beam based on Bayesian theory Sharpening technology,a second-order vector extrapolation acceleration algorithm is proposed in this article.The main contents of this article are as follows:1.The forward-looking scanning radar echo model is established,the echo expression is derived,the ill-conditioned nature of the deconvolution problem is analyzed,and a theoretical basis for subsequent beam sharpening techniques and acceleration algorithms is provided.2.An iterative threshold shrinkage beam sharpening algorithm based on regularization theory is studied.This algorithm establishes a least squares regularization solution model,uses an approximate gradient method to construct an iterative formula,and combines threshold shrinkage to achieve the solution of the deconvolution model,it has features of strong anti-noise;researched the maximum a posteriori beam sharpening algorithm based on Bayesian theory,this algorithm uses the distribution function reasonably to represent the target and the scene's probability model,and realizes the deconvolution by maximizing the a posteriori probability,and it has the characteristics of wide adaptability of scenes.3.The second-order vector extrapolation acceleration method is researched.According to the Taylor series expansion principle,the prediction vector is constructed by using the difference information of the previous two historical vectors,and the iterative path of the prediction vector is modified to achieve iterative acceleration.The conventional methods are compared and applied to iterative threshold shrinkage and maximum posterior beam sharpening algorithm.Second-order vector extrapolation accelerate iterative threshold contraction and second-order vector extrapolation accelerate maximum posterior beam sharpening algorithms are proposed.Under the condition that the beam sharpening performance will not be lost,the iterative convergence speed of the algorithm is effectively improved.All methods in this thesis have been verified by MATLAB simulation and measured data.The results show that the above method can effectively achieve acceleration of different beam sharpening techniques,thereby quickly obtaining super-resolution images of the forward-looking area.