Research On The Application Of Hidden Markov Model In Radar Target Recognition

Radar is an all-weather sensor,which has important application value in military and civil aspects.Radar target recognition is widely used in the national economy,space technology and national defense,and more and more attention has been paid to the countries all over the world.Feature extracting and targets recognition technology based on dynamic RCS series of missile are researched in this project.The research work includes these four content:Firstly,the RCS sequence of radar target is determined by the motion and scattering characteristics of target.According to analyzing the characteristics of RCS variation of radar target,find the difference of RCS sequence variation of different types of target,and lay the foundation for feature extraction.Secondly,the variational feature is extracted using Hidden Markov Model(HHM).The variational characteristic of RCS series of missile is characterized using Hidden Markov Model,and the variational features of midcourse targets include warhead,bait and debris are extracted.Then targets recognition is implemented based on different features of RCS.Thirdly,radar targets recognition and validation using variational feature.The characteristics analysis and feature extraction of dynamic RCS series of missile are implemented based on abundant target property metrical data from the shooting range,the results of validation are well-pleasing.Fourth,Improved implicit Markov model modeling.For the fixed defects of classical hidden Markov model parameters(state number),an adaptive hidden Markov model is established,so that the state number of the model can automatically adapt to the complexity of the signal structure to be modeled,and the adaptive Hidden Markov model is used to characterize the RCS of radar targets,and the classification and recognition of the target are realized.Classification recognition is effective.In this paper,the radar target recognition prototype system based on RCS measurement data is established,which lays a foundation for the target recognition approach to engineering application.