| In recent years,with the rapid development of digital signal processing technology,computer science technology and integrated circuit technology,various new system radar technologies have achieved revolutionary technological breakthroughs.Among them,multifunctional phased array radar has achieved particularly eye-catching development of.Multifunctional phased array radar is currently a common radar system in the battlefield.Multifunctional phased array radar has the characteristics of fast beam scanning,rapid wave position,complex waveform,power control,etc.,which leads to the detection of multifunctional phased array radar facing the problem of small samples and brings challenges to traditional radar working pattern recognition.In the complex electromagnetic environment,how to realize the analysis and intelligent recognition of the working mode of the multi-functional phased array radar with small samples has become a key problem to be solved urgently in modern electronic warfare.This thesis focuses on the intelligent recognition of multi-functional phased array radar track while search(TWS),track and search(TAS),multiple target track(MTT),and single target track(STT)modes in a small sample environment.The main work and contributions of the thesis are as follows:1.Based on the structure of the multifunctional phased array system,the time domain,frequency domain,and space domain characteristics of the track while search,track and search,multi-target track,and single-target track modes are analyzed and the modeling is completed.Summarizes the shortcomings of the existing typical multifunctional phased array radar working mode recognition method for small sample environment.2.Aiming at the problem of multifunctional phased array radar working pattern recognition in a small sample environment,the expansion performance of the conditional generative adversarial network(CGAN),the deep convolution generation adversarial network(DCGAN),and the normal generation countermeasure network(GAN)are analyzed.The impact and sensitivity to the small sample size,clarify the advantages of CGAN;analyze the influence of Alexnet,Letet,and Resnet networks on the intelligent recognition performance of multifunctional phased array working mode,and clarify the advantages of Resnet network.An intelligent recognition method for working mode of small sample multifunctional phased array radar based on CGAN and Resnet is proposed.3.Propose the intelligent recognition method of small sample multifunctional phased array radar working mode based on deep transfer learning and deep LSTM respectively.Researched and compared the fully trained VGG16,VGG19,Dense Net121 network based on the model migration method to intelligently recognize the working mode of the small sample multifunctional phased array radar.At the same time,the feasibility of the deep LSTM network to intelligently recognize the working mode of the small sample multifunctional phased array radar is studied. |
