Research On Micro-motion Feature Extraction For Space Targets At Terahertz Region

Terahertz waves usually refer to electromagnetic waves with frequencies between 0.1~10 THz.The terahertz band lies between the millimeter wave and infrared,which is a transitional band from electronics to photonics.Compared with the traditional microwave/millimeter wave radar systems,the terahertz radar is a main trend in the field of target detection because of its unique advantages of high resolution and Doppler sensitivity,etc.In this paper,in view of ballistic missile defense and high-resolution space investigation,we systemically studied the micro-motion feature extraction for space targets at terahertz region.Specifically,this paper mainly includes the following aspects:The background and significance of this paper were provided in Chapter I.In addition,the current situation and development trend of the terahertz technologies and micro-motion feature extraction in the terahertz band were introduced,especially the main technical approaches and problems.The main contents and the structure of this paper were briefly introduced in the end.Chapter II analyzed the characteristics of micro-motion targets in the terahertz band.The characteristics mainly include micro-Doppler ambiguity,sliding scattering center and rough surface scattering.In addition,the experimental condition of this paper was introduced,especially the terahertz radar systems,the micro-motion simulator,the translation simulator and other accessory equipment.In general,Chapter II is the foundation of this paper.The micro-motion parameter estimation methods under non-ideal conditions in the terahertz band were proposed in Chapter III.The non-ideal conditions refer to the micro-Doppler ambiguity and tiny micro-motion interference problems due to the Doppler sensitivity of terahertz waves.Aiming at the micro-Doppler ambiguity problem,we proposed a micro-Doppler ambiguity resolution method based on intra-pulse interference algorithm.The essence of the algorithm is reducing the equivalent carrier frequency by interference processing of the sample points during the band.The algorithm can estimate the micro-motion parameters accurately.In view of the tiny micro-motion interference,a method based on time-frequency domain filtering was proposed.The method effectively combines the traditional means such as time-frequency analysis,time-frequency domain filtering and Radon transform.It can realize the high-precision parameter estimation of micro-motion targets,as well as the estimation and compensation of the tiny micro-motion interference.Micro-motion parameter estimation methods based on micro-scale features in the terahertz band were studied in Chapter IV.The micro-scale features mentioned here mainly include the size micro-scale feature and the motion micro-scale feature.It can also be called the roughness and the tiny micro-motion of the target from the feature extraction point of view.For micro-motion targets with rough surface,we analyzed the scattering characteristic in the terahertz band,and proposed a micro-motion parameter estimation method based on the peaks in the time-frequency domain.The method can estimate the micro-motion parameters and the size of the rough surface target.As for tiny micro-motion targets,we proposed a method based on the phase-derived-range technology and the empirical mode decomposition algorithm.The phase-derived-range technology is used for measurement of the tiny displacement,and the empirical mode decomposition algorithm is for decomposing and denoising.All the methods were verified by experiments,and the performance analyses were provided.High-resolution imaging of micro-motion targets with terahertz radar systems were studied in Chapter V.Firstly,we proposed a high-resolution/high-frame-rate imaging method for micro-motion targets based on the micro-motion angle,and obtained high-precision micro-motion parameters according to the imaging results.The method was verified by experiments on a precession warhead model.Then we adopted the convolution back-projection algorithm in imaging of micro-motion targets with rough surface.The good imaging results confirmed the effectiveness of the algorithm.In the end,the effect of the high frequency vibration of the platform or the targets on imaging was analyzed according to the specialty of terahertz,and methods based on phase gradient autofocus algorithm and dominant scatterers were provided.In Chapter VI,we studied the translation compensation of micro-motion targets.For targets with micro-motion and translation simultaneously,translation compensation should be achieved before high-precision parameter estimation and high-resolution imaging.However,translation compensation of micro-motion targets is more complicated due to the existence of the micro-motion.For micro-motion targets with a low translation velocity,we proposed a method based on polynomial fitting without regard for the time-varying of the reference distance.For micro-motion targets with a high translation velocity,methods based on the secondary compensation and the multi-layer perceptron were put forward separately.All the methods were verified by detailed theoretical derivation and experiments.The performance curves were given after the introduction of the methods.Finally,the whole work and the main innovation points were summarized in Chapter VII,and some more research aspects and potential application areas were put forward.