Signatures Extraction And Parameters Estimation Of Spatial Multi-target Based On Micro-doppler

At present, target recognition based on Micro-Doppler signatures becomes a new focus and is considered as one of the most potential techniques in radar target recognition field. With the research object of spatial multi-target, this thesis extracts the Micro-motion signatures and estimates Micro-motion parameters of multi-target based on narrow-band radar echo model of space multi-target and multi-component Micro-motion. Then, Micro-Doppler simulation system of narrow-band radar is developed.1. Base on Micro-Doppler effect, the Micro-motion model of space targets, space location model and echo model of multi-target are established. Through researching the shielding effect of scattering point, the attitude angle of Micro-motion is derived. The shielding effect is verified by a series of experiments conducted base on measured data of microwave darkroom.2. For time-frequency map with noise, a method by setting a threshold to reduce noise is proposed, and it can effectively filter noise. The viterbi algorithm is in-depth studied. The algorithm can extract the instantaneous frequency of each component from multi-target and multi-component time-frequency map and separate time-frequency signatures of corresponding component. Aim at overcoming the drawback of viterbi algorithm, local maximum value method which can effectively extract the instantaneous frequency of having discontinuous point in time-frequency map is proposed.3. Through analyzing the Micro-Doppler of the space cone targets, some conclusions are obtained: the function form of cone precession’s Micro-Doppler is harmonic function, and nutation’s Micro-Doppler is not harmonic function, but is a periodic function.4. Based on the instantaneous frequency by local maximum value method obtaining, a period estimation method by accumulating all components is proposed. The method can effectively estimate period of single-target, and roughly estimate period of multi-target. Based on the instantaneous frequency by viterbi algorithm obtaining, the harmonic analysis method which can estimate the period of spin and coning of precession target is proposed. Because this method has estimated error, a compensating method of error is provided. The method can effectively compensate estimated error of harmonic analysis method, when time duration of accumulated echo is more than the period of Micro-motion.5. As Micro-Doppler is sinusoid(a special case of harmonic function), the parametric estimation methods based on sinusoidal function or polynomial function fitting are studied. The two methods can effectively estimate the period, range and phase. As Micro-Doppler is harmonic function, a method based on harmonic function fitting is proposed. The method can effectively estimate the range and phase, but requires harmonic analysis method to estimate period of micro-motion.6. Through designing and analyzing Micro-Doppler simulation system of narrow band radar, the simulation system is implemented by Matlab GUI(Graphical User Interface). This simulation system not only can set parameters of scene and Micromotion on line, but also can check algorithms of signatures extraction and parameters estimation.