Ballistic Midcourse Target Recognition Technique Based On Narrowband Radar Information

Taking full advantage of narrowband ra dar information, the m idcourse target recognition and feature extrac tion technique based on narr owband information is of great significance to the ballistic missile counterwork in midcourse. This dissertation focuses on the problem of midcourse target recognition, serving for the counterwork of ballistic missile defense. The main re search efforts include the active decoy discrimination technique based on orbital ch aracteristic, the period feature extraction technique of precession targets based on narrowband radar cross section(R CS) sequence, and the im aging technique of midcourse precession targets based on time-frequency domain.Chapter 1 illustrates the background and si gnificance of th is research subject, and introduces the present status of research on the counterwork of ballistic m issile. Then the development of midcourse target recognition and feature extraction technique based on narrowband radar infor mation are reviewed and summarized, after which the m ain content of this thesis is presented briefly.Chapter 2 focuses on the basic theory of ba llistic target tracking. The principle of tracking and filtering is introduced firstly, then the mathematic model and essentials of ballistic target tracking based on dynamics model and kinematics model are analyzed. Afterwards, from the two aspects of the prediction performance of state equation and the linearization of filtering process, two novel tr acking algorithms for m idcourse targets are proposed, respectively. One tracking algorithm is based on the earth’s gravity filed and local uniform acceleration m odel, the other is under th e radar vertical coordinate system and based on transform observations. The proposed methods possess the strongpoints of both dynam ics model and kinematics model, i.e., not only have the similar performance as dynamics model, but also have the linearization configuration as kinematics model.Chapter 3 s tudies the active d ecoy discrimination technique based o n the orb ital characteristics of ballistic targets in m idcourse. Firstly, the tra jectory characteristic quantities of ballistic targets and active decoys, such as momentum moment, periapsis vector, and orbital elem ents, are analyz ed thoroughly. Then, an active decoy discrimination method based on the tim e invariant characteristic of orbital elem ents is put forward, m eanwhile a tim e-invariance decision rule is given, and the precise estimate of the initia l observation data is a chieved based o n the princ iple of orbital reversibility. Simulation experiments under the situation of three typical station arrangements verify the validity of the proposed active decoy discrimination method.Chapter 4 researches into the period f eature extraction technique of midcourse precession targets based on narrow band RCS se quence. First the RCS data of so me typical midcourse targets are obtained by el ectromagnetic computing software, and the corresponding dynamic RCS sequence is ge nerated via spline fitting. Then, the traditional fearture extraction method of RCS is presented briefly, and the limitations of traditional method when used for the period feature extraction of precession targets are pointed out. Afterwards, in light of the definition and property of period, two new period estimation methods are proposed, which are respectively based on the similarity measurement and the nonparam etric statistics of grouped da ta. In addition, the period decision rule is also given. Sim ulation results indicate that both the proposed m ethods can overcome the limitations of traditional method effectively.Chapter 5 studies the narrowband im aging technique of midcourse precession targets based on tim e-frequency domain. The trad itional narrowband echo modeling m ethod based on ideal scattering center, the relevant time-frequency analysis technique, and the corresponding narrowband imaging method, are introduced firstly. Through simulation analysis, the deficiency of traditional imaging method when nonideal scattering phenomenon occurs is revealed. Ai ming at this problem, a narrowband echo model based on nonideal scattering center is estab lished for m idcourse precession tar gets, following which a novel imaing method based on time-frequency domain is put forward. The main idea of the proposed m ethod is to decompose the echoed signal by em pirical mode decomposition(EMD) before tim e-frequency transformation, and then perfor m imaging operation in tim e-frequency domain on every intrinsic m ode. Finally, the correctness of the established m odel and the validity of the proposed imaging m ethod are verified by m easured data and precession experiment in anechoic cham ber, respectively.Chapter 6 summarizes the research work and main innovations, and points out the future work to be researched.