Research On Radar Target Recognition Based On Dynamic RCS

Radar target recognition is an important development direction of modern radar research.At present,the research of radar target recognition has made great progress in the aspect of wideband high-resolution radar,and it has more practical significance for the research of conventional radar target recognition.Radar Cross Section(RCS)is the most important amplitude characteristic of the target’s electromagnetic scattering characteristics.In this thesis,Aiming at the problem of aircraft target classification,a the method of dynamic RCS identification using RBF network is proposed.The main work is as follows:Firstly,the basic theory of radar cross section and the RCS solution method of radar target are studied.Considering that the measured dynamic RCS of the target is difficult to obtain,the method for calculating the dynamic RCS of the target is studied.In this thesis,the typical aircraft target is taken as the object,and its CAD model is built and the static RCS value of the target in the whole airspace is calculated.Then,the conversion rules between radar coordinate system and airframe coordinate system are established,and the RCS characteristics of dynamic targets are analyzed with MATLAB software to obtain the target RCS time series.Secondly,in order to improve the effect of radar target feature recognition,the target feature extraction method based on RCS time series is studied.The importance of each eigenvalue in the target RCS statistical feature is analyzed.The location feature,distribution feature and dispersion feature of each target RCS time series are extracted.The discrete wavelet energy feature is extracted by wavelet decomposition and reconstruction of RCS time series.The classification separability measure is introduced,and the validity of the extracted features is preliminarily verified by quantitative analysis of the separability of the fea tures under each target set.Finally,the classification mechanism of RBF neural network is studied deeply,and the RBF neural network based on RCS time series identification is designed.The target RCS time series samples of five kinds of aircraft are use d as input vectors for training and testing.BP neural network algorithm and Knearest-neighbor algorithm are used to design the classifier respectively.By comparing with the two typical algorithms,the recognition performance of RBF neural network based on RCS time series proposed in this paper is evaluated.The simulation results show that the classification effect of this method is better.At the same time,it provides a new idea for the classification of dynamic RCS features,which is of great significance for the development of radar target recognition.