Research On Radar Detection Method Of Low-altitude Micro Target

In recent years,it is the focus of the research to strengthen radar detection capability of the low-altitude targets,specially the micro unmanned aerial vehicle(UAV).Due to the low flight height and small radar cross-section(RCS)of the UAV,the traditional Doppler methods are unable to achieve effective detection of UAV.The thesis researches the problems of identifying the UAV,and the main studies and contributions can be summarized as follows:Firstly,the RCS of common UAV are measured in the anechoic chamber.From the experimental results,the RCS of UAV is similar to the RCS of birds.So the UAV can not be effectively identified with the traditional radar detection methods.We can identify the UAV effectively by detecting the micro-motion of the UAV rotor.Then we establish the micro-motion model,radar echo model and micro-Doppler modulation model of the UAV rotor.Moreover,according to the actual scene,the GTD model is adopted.We propose that the radar echo signal of the UAV rotor is the sinusoidal amplitude modulation-sinusoidal frequency modulation signal.Secondly,the time-frequency distribution methods are used to extract the micro-Doppler spectrum of the UAV rotor based on the above model.Due to the low time-frequency resolution,cross interference and other deficiencies of the traditional time-frequency analysis tools(Short-time Fourier transform,Wigner-Ville distribution and so on),we propose the S-transform with the improved window,L-Wigner distribution and S-class distribution.Then we analyze the performance of the above three methods by theory,simulation and experiment.Finally,we study the separation and parameter estimation of the micro-Doppler signal.In the actual radar detection,the platform returns are aliasing.Because of the strong body echo,the UAV rotor echoes are concealed easily.On the time-frequency distribution,the micro-motion echo of UAV rotor is immersed,which increases the difficulty of analyzing the micro-motion characteristics greatly.So we need to separate the body echo and micro-Doppler signal.In order to identify the UAV further,we need to estimate the parameters of micro-Doppler signal curve.We adopt the Hough transform to estimate the parameters.To increase the calculation efficiency,the random Hough transform is proposed.The adopted method can estimate the micro-Doppler parameters effectively and accurately even in the strong noise conditions.From the simulation and experimental results,the proposed S-distribution method can extract the features of micro-Doppler signals in strong noise environment,and can strengthen the weak-component signal in micro-Doppler signals with great time-frequency resolution.The proposed separation method based on Fourier transform can effectively partition the micro-Doppler signal from the main body echo.The adopted random Hough transform method based on time-frequency ridge can effectively estimate the parameters of micro-Doppler signals in strong noise environment.Compared with the traditional Hough transform,the proposed method has the advantages of higher accuracy,faster running speed and smaller memory footprint.The methods proposed in this paper provides some reference for the detection and identification of UAV.