Research On 3D Attitude Parameter Estimation Method Of Ballistic Target

Ballistic target recognition is the key technology of the ballistic missile defense system,and radar plays an irreplaceable role in ballistic target recognition for it is the core equipment in the system.At the present stage,the similarity of the shape and electromagnetic characteristics between the decoy and the warhead becomes increasingly higher.In this situation,three-dimensional(3D)attitude feature based recognition methods have aroused general interest.The 3D attitude of the ballistic target reflects the movement of different types of targets in the midcourse,and its 3D attitude parameters contain the micro-motion information of the target,which is the basis for ballistic target identification and interception.Therefore,the problem of 3D attitude parameter estimation of ballistic targets has important research significance.This dissertation mainly focuses on the 3D attitude parameter estimation of ballistic targets.The problems in 3D attitude parameter estimation are studied based on HRRP sequences and micro-Doppler.The main work is summarized as follows:1.The high-precision distance estimation problem of the scattering centers based on the wideband radar echo is studied.Firstly,a short-term coherent accumulation model suitable for the wideband radar echo of ballistic targets is established,and the influence of off-grid error on high-precision distance estimation is analyzed.Subsequently,a high-precision distance estimation method based on two-dimensional root sparse Bayesian learning is proposed.The grid position is used as a variable parameter,and the backscatter coefficients and radial ranges of the scattering centers are estimated via the sparse Bayesian learning method and the polynomial rooting method.The simulation results show that the proposed method can obtain the high-precision distances of the scattering centers under the condition of low SNR.2.The 3D attitude estimation problem of the blunt-nosed chamfered cone target is studied.Firstly,the scattering centers of the ballistic target are divided into stable scattering centers and sliding scattering centers according to the scattering characteristics.The radial distance、changes of the two types of scattering centers under micro-motion conditions are studied and verified with electromagnetic simulation data.Subsequently,a more detailed blunt-nosed chamfered cone model is established for the warhead target with a spherical cap and chamfered structure.The scattering centers on the spherical cap and chamfer are modeled as the migrating scattering centers whose positions change with the radar line of sight angle,and the micro-motion equation is re-established.The correctness of the blunt-nosed chamfered cone model is verified by the electromagnetic simulation data.Finally,a 3D attitude parameter estimation method for the blunt-nosed chamfered cone target based on the high-resolution range profile(HRRP)sequences is proposed.In this method the optimization function is constructed based on the radial distance micro-motion equations of the top and bottom scattering centers,and the 3D attitude parameters of the target are estimated by a nonlinear optimization method.3.The problem of 3D attitude parameter estimation of ballistic targets based on multi-station HRRP sequences is studied.Firstly,the multi-station observation models of stable scattering centers and sliding scattering centers are established.Subsequently,a multi-station 3D attitude parameter estimation method based on the strategy of "reconstruction before association" is proposed.In the reconstruction step,the range data are extracted from the wideband echo of each radar in the radar network and are associated with the Kalman filter and the nearest neighbor standard filter method.Then,the biased scattering center coordinates and radar line of sight(LOS)unit vectors are calculated from the range sequences by the factorization method based on the rigid body invariance property.In the association step,the multi-station data are associated with the reconstructed scattering center coordinates,then the coordinates and the 3D attitude parameters in the radar coordinate system are estimated by the relationship between the associated radar LOSs and the real radar LOSs.The simulation results show that the method in this chapter can estimate the 3D attitude parameters correctly of the target under the condition of micro-motion model mismatch.4.The problem of 3D attitude parameter estimation of ballistic targets based on microDoppler is studied.Firstly,the micro-Doppler characteristics of several common warhead models are analyzed,and the general form is given.Then,the narrowband echo of the ballistic target is modeled by the time-varying autoregressive model(TVAR)model.The TVAR model coefficients and micro-Doppler frequencies are estimated via the sparse Bayesian learning fast algorithm based on the sparsity of the coefficients.The fast algorithm avoids matrix inversion,which improve its calculation speed under the condition of low SNR greatly.Subsequently,a micro-Doppler frequency association algorithm based on the frequency-frequency change rate plane is proposed.Under the condition that the frequencyfrequency change rate curves of different scattering centers are disjoint,the micro-Doppler frequency and frequency change rate are estimated by TVAR model coefficients and associated on the frequency-frequency change rate plane.Finally,the 3D attitude parameters of the warhead target are estimated by trajectory information and the micro-Doppler characteristics of different scattering centers.