Research On Azimuth-Pitch Two-Dimensional Super-Resolution Imaging Method For Scanning Radar

Scanning radar can quickly obtain the three-dimensional spatial coordinate information of targets in the wide-area search area by scanning two-dimensional beams in azimuth-pitch,which plays an important role in applications such as formation target recognition,weather monitoring,and path navigation.However,limited by the size of the radar antenna,the azimuth-pitch two-dimensional angular resolution is low,and it is hard to achieve angular resolution of multiple targets at long distances in space,which can easily lead to misjudgment of the number of targets.Super-resolution imaging technology does not require software and hardware modification,and can obtain angular resolution beyond the antenna beam width only through signal processing methods,which is an important way to improve the angular resolution capability of scanning radars.However,the existing super-resolution methods are mainly aimed at the onedimensional beam scanning mode,and there is no ready-made solution for the azimuthpitch jointed super-resolution problem.Aiming at this problem,this thesis focuses on the azimuth-pitch two-dimensional imaging model,two-dimensional super-resolution method and fast super-resolution method.The main work is as follows:1.The azimuth-pitch two-dimensional imaging geometric model of the scanning radar was studied,the distance history expression between the radar and the target was derived,and the azimuth-pitch two dimensional super-resolution echo model was established.The system parameter constraints affecting the super-resolution performance were analyzed and derived,which laid the foundation for the subsequent twodimensional super-resolution imaging method.2.A L1-L2 mixed-norm two-dimensional regularization super-resolution method is proposed.Using the matrix Kronecker product and vector columnarization properties,the two-dimensional deconvolution is converted into one dimensional form,and then by adding regularization prior constraint information,the azimuth-pitch two-dimensional multi-target resolution problem under low signal-to-noise ratio is solved.3.A fast sparse sparse iterative minimization two-dimensional super-resolution method is proposed.By using conjugate gradient linear iterative optimization,which avoids the high-dimensional matrix inversion operation in iterative deconvolution.The high computational complexity of the 2D deconvolution super-resolution method after vectorization is solved.The effectiveness of the above models and methods have been verified in simulation experiments.The results show that the fast two-dimensional super-resolution method proposed in this thesis not only improves the beam sharpening significantly,but also improves the computational efficiency by about 20 times.