Characteristics Analysis And Feature Extraction Technique Of Warhead Target

The recognition of the Middle ballistic target is a key technology in the ballisticmissile defensive system. It draws great attentions to the researchers to quicklyrecognize the missile from the target complex with baits and debris. Feature extractionis an important means for target identification. To the warhead target, the scatteringcharacteristics, geometric characteristics and micro-mtion characteristics are the mainfeatures to separate with baits and debris. This dissertation researchs the problem ofwarhead target feature extraction from the electromagnetic scattering characteristics andmicro-motion characteristics of warhead target. The main sduties can be summarized asfollows:In chapter II, the electromagnetic scattering of the warhead is studied, and aparametric model of its non-uniformly distributed scattering center is proposed. Firstly,a deep research on attributed scattering center model and its parameter estimationmethods is proceeded, and the problem that the attributed scattering center model cannot accurately describe the azimuth distribution of non-uniformly distributed scatteringcenter is pointed out with an in-depth analysis on the model. Aiming to solve thisproblem, a new parametric model is deduced with a2D Fourier transform on the ISARimage data of the non-uniformly distributed scattering center. And then, a parameterestimation method is proposed to fit the new model, and a simulation is conducted tovalidate the correctness of the model.In chapter III, a research on the motion characteristics of the warhead target in themiddle is proceeded. Firstly, a research on the traits of changement of the warheadtarget pose angles to the radar line of sight is proposed. Then the echo simulationmethod and the scattering coefficient extraction are introduced. Finally, combined withsimulation data, the characteristics of the target length in the range-picture sequence andthe relation between the cross-range resolution of ISAR images and the pose angles areanalyzed.In chapter IV, the new feature extraction method based on the crosse-range scalingof ISAR image is introduced in detail. Firstly, the principle of ISAR image scaling isrecommended briefly. On the basis of the principle and the pose angles caused by targetprecession, the cross-range resolution and the clearness of ISAR image cannot reach therequirement together. In order to solve this contradiction, the new method is proposed.This method employs data resample to remedy the non-uniform variation in the poseangles caused by target precession. This method takes full advantage of the variation inpose angles caused by the precession of the ballistic target. It has high scaling accuracyand stable performance as well as requiring short observation time and smallcomputational efforts. Finally, Experimental results based on the ballistic target motion model and simulated electromagnetic data are provided to verify the validity of themethod.At the end of this dissertation, a systematic summary of the main content ispresented. The problems that not resolve in this peper and the research directions for thenext step are pointed out.