Study Of Radar Detection Methods Of High-speed Moving Targets

In modern warfare,high-speeds moving targets have posed a severe challenge to the detection ability of radar.In general,the long time accumulation technique is an effective way to improve the detection ability of radar to high-speed moving target.However,because of the complexity of high-speed target movement,it is easy to produce distance migration and Doppler frequency expansion in the long time accumulation process,which leads to the accumulation of energy diffusion and the lower detection performance.In this thesis,the problem of distance migration and Doppler frequency expansion is studied mainly around pulse radar system.Firstly,the high speed moving target echo signal is modeled and the relationship between the distance migration and the Doppler frequency expansion problem and the target motion parameters and radar parameters is analyzed.Then,the effect of this problem on accumulation gain of traditional accumulation method is quantitatively analyzed.The simulation results show that the degree of distance migration and Doppler frequency expansion will influence the accumulation gain.The greater the degree of movement,the more serious the accumulated loss.Aiming at the problem of distance migration,an improved method is proposed to improve the performance of the algorithm based on the traditional Hough transform detection algorithm.The method firstly improves the SNR of the echo signal by the non-coherent accumulation.Then,the data points over the first threshold are treated with binary processing to eliminate the energy difference between weak and strong targets,which can effectively prevent the weak target loss in multi-target detection.In addition,the noise points can be removed and the algorithm operation amount can be reduced effectively by clustering processing after binary processing.In view of the problem of Doppler ambiguity in first order Keystone transform and blind speed sidelobe in discrete Radon-Fourier transform,the corresponding solution is given,and its validity is verified by simulation.At last,the improved Hough transform processing method,first order Keystone transform and Radon-fourier transform are analyzed and compared from three aspects of multi-target detection,detection performance and computational quantity,and the advantages and disadvantages of each method are summarized.Aiming at the problem of Doppler frequency expansion,based on DPT and Fr FT algorithm,a joint parameter estimation method based on DPT and Zoom-Fr FT is proposed.The algorithm first obtains the rough measurement value of the parameters through the DPT algorithm,and determines the range of the value of the parameters.Then the parameters are further refined by Zoom-Fr FT algorithm.The relationship between the target parameters and the Zoom-Fr FT algorithm parameters is used to limit the range and resolution of the algorithm parameters,further reducing the computation of the Zoom-Fr FT algorithm and having higher parameter estimation precision.Simulation analysis shows that the noise threshold of this algorithm is about 3d B lower than that of the DPT algorithm,which is more suitable for the parameter estimation in the case of the low SNR.In addition,in view of the problem of the parameter estimation fuzzy in LVT algorithm,an improved LVT algorithm is given,and the validity is verified by simulation.