Research On Radar Emitter Recognition Method Based On Support Vector Machine

As an important equipment of Electronic Warfare,radar is massively equipped to the warring sides,which not only improves the reconnaissance capability,but also leads to more and more complicated electromagnetic environment in the battlefield.How to quickly and accurately identify the radar radiation sources of the enemy and us in the complex electromagnetic environment is an important research direction of electronic warfare.The classification recognition and multi-sensor fusion recognition of radar emitter signal are studied in this thesis.The fireworks algorithm and evidence theory are applied to the recognition of radar emitter to realize the effective recognition of radar emitter.The main work of this paper is as follows:Firstly,the basic theory of SVM is studied in this thesis,Through the simulation experiment,the classification effect of SVM classification models with different kernel types and kernel parameters is compared and analyzed,leading to the problem of SVM parameter optimization;Secondly,the advantages and disadvantages of the traditional fireworks algorithm are studied in this thesis,an improved fireworks algorithm(ifwa)is proposed based on the gamma distribution and nonlinear mutation,and the performance of the improved algorithm tested through the standard test function,the simulation results show that the improved algorithm has strong optimization ability;based on the analysis of the characteristic parameters of radar emitter,a radar emitter recognition model based on the improved fireworks algorithm is proposed,and the effectiveness of the algorithm is verified by simulation test;Finally,a decision-making level radar emitter fusion recognition method based on evidence theory is studied in this thesis,based on the analysis of multi-source fusion recognition,evidence theory is applied to radar emitter fusion recognition,in order to solve the phenomenon of "one veto" in decision-making level fusion in high conflict situations,an adaptive effective factor fusion method which does not rely on prior information is proposed,the method modifies high conflict evidence to achieve effective fusion in high conflict situations,the improved method is applied to decision-making radar recognition,which effectively improves the radar recognition effect in high conflict situations.