Research On Dual-channel Millimeter-wave Alarm Receiver And Algorithm

This paper develops a set of millimeter-wave warning receiver with millimeter-wave detection or guidance signal which can effectively sensitive and identify incoming munitions,The design scheme of millimeter-wave warning receiver is put forward,the principle and algorithm of millimeter-wave alarm is studied,the integrated design of hardware and software of the alarm system is completed.and the field experiments is completed.Satisfactory results are obtained.The details are as follows:Firstly,based on the engineering requirements of the actual project,the design scheme of millimeter-wave alarm is put forward.The system of millimeter-wave alarm is divided into receiver hardware subsystem and signal processing subsystem,and the corresponding subsystem is demonstrated.Secondly,the composition of millimeter-wave alarm system and the principle of each module are analyzed.The circularly polarized receiving antenna is studied,and the performance of the circular polarized antenna is analyzed and simulated.The function composition and working principle of millimeter-wave receiving front end and geophone circuit are analyzed in detail.Thirdly,a low noise and low frequency circuit with dynamic intelligent adjustment function is designed.The amplitude phase characteristic,bandwidth and gain of low frequency circuit are simulated.At the same time,the embedded acquisition and target recognition circuit which can meet the needs of high-speed acquisition,fast transmission and recognition,and the high-efficiency power conversion circuit are designed.Fourthly,the system software workflow of the millimeter-wave alarm is discussed,and the initialization module,self-test module and data acquisition module of the system software are described in detail.The software design of the data processing module,target recognition module and adaptive adjustment module of the alarm is completed,and the problem of false alarm and missed alarm is solved emphatically.Fifthly,the wavelet transform mode maximum algorithm is proposed to denoise the received signal of the millimeter-wave alarms,the Gabor atomic transform algorithm is used to extract signal features.The recognition algorithms of Support Vector Machine(SVM)and Relevance Vector Machine(RVM)are studied.The simulation tests of SVM and RVM are carried out from the training time and testing time of the recognition algorithm and the average recognition rate,and their performances are verified.Finally,the field test system of the millimeter-wave alarm and the data analysis software to be used in the experiment are designed.The performance test and the field experiment of the alarm are completed.The data signals in different environments is analyzed,and the experimental data acquisition is identified and classified by using two algorithms of support vector machine(SVM)and correlation vector machine(RVM).The possibility of the application of the alarm in the actual battlefield is verified by analyzing the experimental data.