Research On Recognition Of Space Cone-shaped Targets Based On Narrowband Radar Feature

The space cone target recognition occupies an important place in the national defense system, and radar plays an irreplaceable role in space cone target recognition for it is the main detection and tracking device in the system. In recent years, the modulation induced by micro-motion has been used in recognition more and more often, also it is considered as a promising way to solve recognition problem for the forceful reflection of targets’ characteristics. At the present stage, narrowband radar is still widely used in our country, and the advantages such as long tracking distance, clear micro-Doppler signature make it a good candidate for recognition. Based on the these analysis, this dissertation gives research on the space cone target recognition problem based on narrowband radar micro-motion modulation, mainly focusing on the following aspects: the modeling of target dynamic scattering characteristics, the estimation of Doppler frequency, the classification of micro-motion form, the estimation of target’s size and micro-motion parameters. The main contributions of this dissertation are given as follows:1. The characteristics of space cone target’s orbit movement are firstly analyzed, then traditional point-scattering model and equivalent point-scattering model are established to describe different kinds of targets, the difference between the two models in micro-Doppler modulation under different micro-motion forms and occlusion effects are also analyzed, electromagnetic computation data is used to verify the analysis.2. The estimation method of targets’ Doppler frequency and echo compensation problem are discussed. The time-varying autoregressive(TAVR) model is applied to estimate the Doppler frequency. By learning the time-varying parameters the signal’s power spectrum intensity(PSI) can be acquired, then the scattering center’s Doppler frequencies are calculated using the poles of PSI on every moment. The Kalman filter is used to separate and re-associate these discrete frequency components. Then frequency relocation method is applied to solve the ambiguous problem arises in the Doppler frequency estimation results, the envelop of the un-ambiguous frequency is polynomial fitted and the fitting coefficients are used to compensate the target’s translation Doppler frequency, which makes sure the narrowband echo compensation and micro-Doppler frequency estimation can be accomplished at the same time.3. The classification of target micro-motion forms is studied based on smooth cone target which satisfies the equivalent point-scattering model. First the scattering centers’ micro-Doppler frequency variation area on time-frequency(TF) distribution is extracted using energy principle. After that, the TF area differences between spin, precession and nutation are discussed, based on which four characters are extracted from the TF area and the micro-motion forms are classified using the four characters. The influence to classification results induced by different radar parameters is also discussed, which lays the foundation of target size and micro-motion parameter estimation based on precession.4. Precession target size and micro-motion parameter estimation methods based on scattering centers’ micro-Doppler frequencies are proposed. First, aiming at the space cone target with zero-incidence characteristic, the angle variation versus time between precession axis and radar line of sight is calculated by trajectory prediction. Then the location of center of mass is eliminated using bottom scattering center micro-Doppler frequency variation, after which the precession angle and bottom radius are acquired. Finally the target height and location of center of mass are calculated using top scattering center’s micro-Doppler frequency variation.For the scenario where the trajectory cannot be utilized, a parameter estimation method based on micro-Doppler frequency expansion is also proposed. The bottom scattering center micro-Doppler frequency variation is Talyor expanded firstly, then the frequency reconstruction error versus expansion order are calculated, which indicates the demands for accuracy can be satisfied using six order expansion. After that, the relationship between the first six expansion coefficients and target parameters are derived, and the coefficients are calculated by solving linear system of equations built by combing bottom and top scattering center micro-Doppler frequency variation. Finally, target size and micro-motion parameters are acquired using these coefficients.5. A target micro-motion and size parameter estimation method for narrowband radar networks is proposed. A approximate estimation method for radar line of sight and precession angle under the special circumstance that target’s center of mass locates on its bottom is presented firstly, then the top scattering center micro-Doppler frequency components is used to compensate bottom components and the best compensation coefficient is decided by variance of the radar line of sight and precession angle acquired by approximate method. After that, the radar line of sight and precession angle are calculated by the best compensation coefficient, based on which the target height, bottom radius and location of center of mass are also acquired.