Research On ISAR Imaging Method Of Complex Moving Targets Based On Sparse Time-varying Autoregressio

Inverse Synthetic Aperture Radar(ISAR)has the capability of all-day,all-weather,and long-distance work.Due to its high-resolution target imaging capability,it is a very practical technical means in air defense,anti-ship and anti-submarine warfare.However,for air targets with high maneuvering flight,such as enemy aircraft and missiles,the traditional range-Doppler(RD)algorithm cannot achieve high-resolution ISAR imaging,resulting in the inability to implement effective interception and strike.The research on ISAR imaging of complex moving targets has very important practical application value.This paper mainly focuses on ISAR imaging of complex moving ballistic missile targets,and proposes an ISAR imaging algorithm for complex moving targets based on sparse time-varying autoregression.The specific research contents are as follows:1.Study the micro-Doppler characteristics of space cone targets.Build a ballistic missile target model,analyze the micro-Doppler frequency variation law of the target caused by three kinds of micro-motions: spin,precession and nutation,and use short-time Fourier transform(STFT)and adaptive optimal kernel(AOK)two traditional time-frequency analysis methods to perform micro-Doppler analysis on the target.2.A time-frequency analysis method based on sparse time-varying autoregression(TVAR)is studied.The traditional STFT and AOK time-frequency analysis methods have the problems of low time-frequency resolution and cross-term interference;the least squares(LS)method to solve the TAVR model has the problem that the accuracy of the coefficient solution depends on the singularity of the matrix;based on block sparse Bayes BSBL(BSBL)solves the TVAR model by using the block sparse characteristic of the time-invariant coefficients of the model to solve the uniform block coefficient vector with known block boundaries,and the time-frequency plane is relatively clear,but the algorithm does not consider the non-ideal nature of the actual target.Scattering properties.In practical applications,the block sparse time-invariant coefficients of the TVAR model are unknown to the block boundary.Therefore,this paper proposes an algorithm to solve the forward and backward TVR model based on the cluster structure prior,using the Extended Block Sparse Bayesian Learning(EBSBL)algorithm.The improved algorithm obtains the prior representation of the clustering structure of the time-invariant coefficients,and then appropriately handles the neighborhood hyperparameters to promote the correlation of adjacent sparse coefficients to solve the time-invariant coefficients.In view of the problem of high computational complexity caused by the matrix inversion of the algorithm,the efficient sparse Bayesian learning(ESBL)is used to realize its fast algorithm.The simulation and experimental results show that the proposed algorithm can obtain a clearer time-frequency diagram,has a fast operation speed,and has strong anti-noise performance.3.A distributed ISAR imaging algorithm for complex moving targets based on sparse time-varying autoregression is studied.Due to the Doppler time-variation of the scattered point echoes of complex moving targets,the RD imaging results will have serious azimuthal defocusing problems.Therefore,the proposed range-instantaneous Doppler(R-ID)imaging of the target is performed using the time-frequency distribution of the forward and backward TVAR model based on the cluster structure prior,and the time-frequency analysis is performed on the echo signal of each range unit of the target.Obtain the high-resolution instantaneous Doppler spectrum of the scattering point at each moment,and then obtain the ISAR image of the target.Secondly,the multiple-input multiple-output(MIMO)distributed ISAR imaging technology is combined to observe the target from different angles to improve the azimuth resolution of the imaging.Simulation experiments demonstrate the effectiveness and superiority of the proposed method.