Research Of Long-time Integration Algorithm For High Speed Radar Target Detection

The strategic position and military value of near space are becoming more andmore important. Hypersonic vehicles in near space have a particularly strongpenetration ability, which can effectively break through the existing radar defencesystem and bring a great threat to national security. Thus improving the hypersonicmaneuvering target detection ability of radar on the condition of low signal to noise(SNR) has become an extremely important task and serious mission under such a severesituation. The detection performance of high maneuvering target in extremely complexenvironment can be effectively improved by long-time integration technology of staringradar. Then the threats can be effectively and reliably detected. As a result, thetechnology becomes a hot research at home and abroad.This thesis focuses on the high speed target detection problem, and specificallydiscusses and studies the long-time integration technology of staring Pulse Doppler (PD)radar. The main contributions are as follows:1. The basic theory of radar signal long-time integration technology is expounded.The definition of “long time” in the integration is given. The problems such as echocoherent in long time are discussed. The problems such as the advantages andinfluences of long-time integration for the “low detectable target” are analyzed. Finally，aiming at radar pulse echo acquisition problem for high-speed target detection,introduced staring mode of the PD radar system.2. Aiming at the detection problem of near space hypersonic vehicle, the motionmodels and the corresponding radar echo models are established. The motion model ofperiodic trajectory jumping target is discussed first. Combining target motioncharacteristics with actual detection situation of monostatic radar, the motion models ofradial uniform, oblique uniform linear and radial uniformly accelerated rectilinear targetare reasonably constructed. Then radar echo models of staring PD radar are establishedaccording to the three types targets. The characteristics of radar echo are analyzed, andthe applicable conditions and occasions of each model are pointed out.3. Aiming at the crossing rang-cell and crossing Doppler-cell problem of high-speed maneuvering target detection of staring PD radar, three kinds of long-timecoherent integration algorithms are studied, which are long-time coherent integrationalgorithm based on Generalized second order Keystone transform, long-time coherentintegration algorithm based on Keystone transform and MFRFT/MDCFT parameterestimation and long-time coherent integration algorithm based on focus-before-detect(FBD) three-dimensional Generalized Radon-Fourier transform (GRFT). Thesealgorithms not only correct range migration but also effectively solve the Doppler highorder compensation problem, which result in the great improvement of integration gain.The above work has been verified by computer simulation. For near spacehypersonic vehicles, the methods in this thesis can effectively solve the long-timeintegration problem in case of crossing range-cell and crossing Doppler-cell on thecondition of low SNR, which greatly enhance the high-speed maneuvering targetdetection performance of staring PD radar.