Three-Dimensional Scattering Centers Reconstruction And Recognition For Micro-motion Targets Based On Wideband Radar

With the development of high resolution radar technology and the enhancement of dependence on the space in modern war,the detection and recognition of space targets becomes a research focus recently.In this thesis,we study on the algorithm of reconstruction and recognition for three-dimension scattering centers of micro-motion targets based on wideband radar.The main research contents and innovations:In the first part,three-dimensional motion model of space targets is established.Then,derive the azimuth angle and elevation angle formulations of three micro-motion models including rotation,precession and small-angle micro-rotation separately.And simulate one-dimension range profiles of space targets including warhead,cabin,decoy of cone and decoy of sphere based on electromagnetic simulation data.Finally,we analysis the effect of two-dimensional and three-dimensional rotation on the phase histories of the scattering centers and establish the basic framework process of the imaging / reconstruction of the micro-motion targets.In the second part,the extraction algorithm of 1-D scattering centers phase history is addressed,which is the precondition of the algorithm of three-dimensional scattering centers reconstruction.Based on the study of existing methods about extraction of phase history,we focus on the performance of the Radon_CPF+NLS algorithm by simulation,and test the impact of various factors on the performance of the algorithm.Through the simulation,we can get some experience thresholds to give advice to extraction of phase history,which provides a theoretical support for subsequent application.In the third part,the existing reconstruction algorithm based on phase history matrix will be improved.First,to improve the accuracy of the reconstruction target,screen correct ones form all phase histories.Then,propose a method to eliminate the impact of translation motion through the analysis of influence of translational motion to the phase history matrix.Finally,design experiment to verify the validity of the improving process,and obtain the focused ISAR image or distorted three-dimensional scattering centers.In the fourth part,explore recognition algorithm processes for distorted three-dimensional scattering centers.Apply moment features extraction and BP neural network method to achieve the purpose of recognition between warhead and decoy of sphere.In the simulation,we get the confusion matrix of the recognition results.The recognition rate of warhead reaches 96%,verifying the validity of the algorithm.