Font Size: a A A

Research On Target Classification Method Of Frequency Modulation Continuous Wave Radar

Posted on:2018-07-10Degree:MasterType:Thesis
Country:ChinaCandidate:J B FengFull Text:PDF
GTID:2348330542450952Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
The introduction of the micro-doppler effect has a milestone significance in the development of radar target recognition technology,and provides a new research idea for radar target recognition.The micro-doppler effect refers to the vibration or rotation of the target parts,resulting in a modulation component that is different from the target.The geometrical structure and motion characteristics of different targets can produce different micro-doppler modulation components.These modulation components provide the basis for radar target classification and recognition.When the parts is moving slightly faster,the echo is dechirped,the micro-doppler information of the echo distributes in the echo of several distance units,and the phenomenon of spectral peak splitting happened.On this occasion,using the narrowband target classification method of conventional pulse-doppler radar will cause some performance loss.In response to this problem,this paper mainly study the target classification method of narrowband FMCW radar.The main work consists of the following sections.Firstly,the processing method of the signal,sawtooth frequency modulated and triangular frequency modulated,is analyzed simply.Based on the method of decoupling in triangular frequency modulated signal,the phenomenon that range-velocity coupling in the target echo is discussed.Secondly,in the narrowband FMCW radar system,this paper analysis the aircrafts’ and vehicles’ characteristics of micro-doppler effect.The echo model of the target inching parts(rotor、track and wheel)under FMCW radar is derived,and the characteristics of the echo are analyzed.When the scope range of micro-doppler,corresponding spectrum,is wider than the range resolution,the splitting of the spectral peaks occurs and the information of micro motion is distributed over a plurality of distance units,which is diffticult to extract the features.When the phenomenon,splitting of the spectral peaks,is obviously in the target echo,the echo of micro-doppler which is distributes in multiple distance unit is processed.Finally,in order to solve the question of feature extraction from the echo of micro-doppler which is distributes in multiple distance unit,this paper proposes a method of fusion feature,two-dimensional image feature and so on,and the effectiveness of various methods is verified by combined simulation data.
Keywords/Search Tags:Micro-Doppler, FMCW, Target Classification, Feature Extraction, Spectral Peak Splitting
PDF Full Text Request
Related items