Airborne Radar Target Recognition Technology Research

This thesis is focused on the study of classification algorithm for radar airborne objectives. Two approaches of recognition are studied. The main creative work includes:a)Signal modeling for Micro-Doppler effect is accomplished. Improved Empirical Mode Decomposition is adopted to extract the features of micro-Doppler signal, and the classification method using correlation matching algorithm is applied. The recognition results of three kinds of simple targets are given .Programming of the classification algorithm is accomplished and the experiment results with simulation data and the real target illustrate the correction and effectiveness of the designed software.b)A novel radar target recognition algorithm capable of on-line training is proposed by applying centre distance ratio into SVM based on High Range Resolution Profile (HRRP).IFFT is adopted to extract target features. Redundant information is reduced by using window smoothing technique, power transformation and normalization. Power averaging is used to improve the extracted features. The classification is performed by centre distance ratio algorithm based on support vector machine. Classification program is compiled and the algorithm is verified by the real targets data and simulation results.