Research On High Speed Multi-target Detection In Space

Compared with traditional narrowband radar,wideband radar has high range resolution and has outstanding advantages in many aspects.For example,wideband radar can obtain more feature information about targets,more accurate measurement of target motion parameters,better target identification ability,better anti-jamming performance and so on.Therefore,wideband radars can cope with the challenges and increasingly complex tasks of radars in various fields in today's new environment.The study of the theory and application of wideband radar signal processing is a hot topic in the current military research and has a bright application prospect.In order to make better use of wideband radar in practical application,the performance impact brought by its large wideband to high-speed target detection and tracking technology urgently needs to be solved.This paper mainly studies the high-speed moving target detection under the wideband radar system.The high-speed moving target echo model is established,focusing on the analysis of the problems of range migration,range curvature and Doppler ambiguity caused by large bandwidth and corresponding solutions.The main research contents and results of this article are summarized as follows:1.The coherent accumulation under wideband conditions is studied.First,a wideband radar high-speed target echo signal model is established,and the causes of range migration and the effects on coherent accumulation performance are analyzed.This paper introduces the principle and two implementations of Keystone transform to correct range migration,and explains the Doppler ambiguity method in keystone transform.The effectiveness of the Keystone transform to correct linear range migration is verified by simulation.2.Considering that Keystone transform can only correct linear range migration caused by speed,a high-speed wideband radar multi-target detection method based on Histogram Probability Multiple Hypothesis Tracking algorithm(H-PMHT)was studied for the range curvature caused by acceleration.Firstly,based on the theoretical basis of H-PMHT algorithm,the simulation of H-PMHT algorithm to track two uniform linear moving targets verifies the feasibility of the algorithm to track multiple targets under the condition of low SNR.Then,the H-PMHT algorithm is applied to the detection of wideband high-speed target,a state transition model is established according to the motion characteristics of the high-speed targets,and the target echo data that has been processed by Keystone transformation and compensated by Doppler ambiguity is taken as the target measurement model.This method uses the distance-time graph of the target motion as the observation histogram of the stationary random process,assigns the total number of observation data of each resolution unit in the figure to the target and noise model,and finally uses the EM algorithm to correlate the data.Quantify to achieve optimal tracking under maximum likelihood.The H-PMHT algorithm under low signal-to-noise ratio is simulated to track two uniformly accelerated moving targets.The tracking error and intensity estimation are analyzed to verify that the method has good tracking effect and low computational complexity.3.The multi-pulse coherent accumulation method of high-speed group target of wideband radar is studied.In addition to the problems of range migration and Doppler ambiguity,high-speed group targets also have two unique problems: 1)The Doppler ambiguity numbers of the radar returns from a single target may be different at the frequency points within the bandwidth;2)For Doppler of multiple targets with similar speed,the number of ambiguity is different.After analysis,the traditional Keystone transformation cannot solve these two problems.Based on the idea of Keystone transform,combined with the characteristics of low-pulse repetition frequency wideband signal natural under sampling,a coherent accumulation method of wideband radar high-speed group targets based on Sparse Bayesian Learning is proposed.This method uses the sparseness of the Doppler domain to establish a sparse representation model of the high-speed target signal,recalibrates each frequency component,and takes the grid mismatch into account.The first-order Taylor expansion is used to modify the dictionary,and the Sparse Bayesian Learning algorithm is selected to reconstruct the target echo signal.It is verified by simulation that the method can overcome the two problems mentioned above,suppress Doppler ambiguity,and realize wideband radar high-speed group target detection and resolution.Comparing the modified Sparse Bayesian Learning algorithm with the traditional Sparse Bayesian Learning algorithm,it is verified that the former has higher resolution.