Research On Target Classification And Parameter Estimation Methods For Near-ground Millimeter-wave Detection

With the requirement of small and medium-sized intelligent munitions such as smart mines to detect target near the ground,related methods of target distance speed parameter estimation and target classification as a platform to the millimeter wave LFMCW radar are studied in the dissertation.mainly include:(1)Aiming at the problem that the ground clutter environment has difficulty in detecting moving targets,an MTD processing scheme based on the second-order MTI cancellation + FFT method is designed to perform ground clutter suppression.The two-dimensional FFT-based differential frequency signal processing method to implement the LFMCW radar range-velocity decoupling method is analyzed.The simulation illustrates that this method causes velocity ambiguity and then affects target distance and speed parameter estimation.In response to the resulting speed ambiguity the variable interval FM sequence waveform and the corresponding phase difference solution,Doppler algorithm are used to solve the problem of speed ambiguity and the accurate estimation of the target distance and speed parameters is achieved.Simulation results verify the effectiveness of the algorithm.A two-dimensional CA-CFAR detector using a two-dimensional “cross” sliding window is used for target detection of two-dimensional spectrum.(2)Scattering center model of wheeled vehicle and caterpillar in optical region,the High Resolution Range Profile of target is obtained by using FFT.Based on the analyses of the sensitivity issues of HRRP,the pre-preprocessing of them is studied.Speed compensation is achieved by the estimation of target speed parameter,then parallel movement sensitivity of HRRP has been solved.The attitude sensitivity of HRRP has been solved by using the sigmoid transform and frame method of HRRP based on correlation coefficient.Features which is described the figure of HRRP is extracted by using the structure feature of target,then target classification is achieved.And,feature which id described the figure of spectrum is extracted by using wave entropy and difference sum modulo value based on the model,then target classification is achieved.The micro-motion parameters of the target and target classification are achieved based on the result of time-frequency analysis by using Short-Time Fourier Transform.(3)A signal processing system for MMW LFMCW radar is designed and implemented and the debugging of its software and hardware are finished.The function of the signal processing algorithm has been verified by analyzing the signal data of car and truck targets in field experiment.