Study On 3D High-resolution Imaging Of Micro-motion Targets

For high resolution radar,micro-motion refers to the small scale,non-uniform motion of the target or some structures on the target along the radar line of sight.By high-resolution three-dimensional(3D)ISAR imaging of micro-motion targets,we can obtain the structure and motion information,which provides significant features for detection,tracking,classification and identification,and plays a vital role in space situation awareness and ballistic missile defense.Due to the complex micro-motion and the induced highly non-stationary echoes,the available ISAR imaging methods become invalid.To solve this problem,this thesis studies high-resolution 3D ISAR imaging of micromotion targets based on trajectory matrix factorization from four aspects,i.e.typical micro-motion modeling,high-resolution 3D imaging of single micro-motion target,high-resolution 3D imaging of group micro-motion targets,and typical micromotion parameter estimation for precession targets.The content of this thesis is summarized as follows:The first part briefly introduces the basic theory of ISAR imaging and derives the geometric models of micro-motion targets.The second part focuses on high-resolution 3D imaging of single micromotion target based on trajectory matrix factorization.To improve the separability of scatterers,the scatterer trajectory association method based on the range-instantaneous-Doppler(RID)image series is proposed.Firstly,the watershed method is applied to extract the two-dimensional coordinates of scatterers in the RID image.Then,the trajectory association is achieved by the minimum Euclidean distance criterion.To refine the trajectory matrix estimation,a method based on spectrum analysis is proposed.Finally,the 3D scatterer distribution of micro-motion targets is obtained by trajectory matrix factorization.Compared with parametric imaging methods,this method has robustness to complex micro-motion and low computational complexity.The validity of the method is proved by the simulation data of precession and nutation cones.The third part focuses on high-resolution 3D imaging of group micro-motion targets.For trajectory matrix reconstruction of group micro-motion targets in complex environment,a trajectory matrix estimation method based on the augmented Lagrangian technique is proposed.To separate the trajectory matrix,a clustering method based on k-means and graph-Lagrangian is proposed.Finally,trajectory matrix factorization is implemented to separated sub-targets to realize 3D imaging of group micro-motion targets.The validity of the algorithm is verified by simulation.The fourth part focuses on micro-motion parameter estimation of precession targets.The proposed method makes full use of the micro-motion information contained in the radar LOS matrix after trajectory matrix decomposition,and derives the coordinate of equivalent radar LOS in the third-dimension.Then,the nonlinear least square and quadratic programming techniques are applied to estimate the Euler rotation matrix and the precession angle.The proposed method is independent of scatterer distribution,and effectively avoids the high computational burden in parameter estimation induced by the coupling between micro-motion parameters and scatterer coordinates.The validity of the method is verified by simulations.