基于机器学习的辐射源行为建模和感知软件设计与实现

Design and Implementation of Emitter Behavior Modeling and Perception Software Based on Machine Learning

雷达辐射源行为建模和感知的主要目的是对复杂电磁环境战场中的雷达信号进行分选和识别。设计了一种基于机器学习的辐射源行为建模和感知软件,聚焦于辐射源目标行为判别。首先利用时频分析协同融合模型进行特征提取,然后建立深度学习模型进行分类识别,获取辐射源参数、扫描特性、调制规律等相关参数信息,对雷达工作模式进行识别,显示辐射源分布态势,为后续的威胁等级计算提供依据。对软件各模块提出了工作原理和仿真结果,结果表明,该方法可以对多个辐射源进行有效识别。

The main goal of radar emitter behavior modeling and perception is to sort and recognize the radar signals in the bat⁃tlefield with complex electromagnetic environment.In this paper,an emitter behavior modeling and perception software based on ma⁃chine learning are designed to focus on emitter target behavior discrimination.Firstly,the time-frequency analysis collaborative fu⁃sion model is used for feature extraction,and then the deep learning model is established for classification and recognition.The emit⁃ter parameters,scanning characteristics,modulation rules and other related parameters are obtained to identify the radar working mode,display the emitter distribution situation,and provide the basis for the subsequent threat level calculation.The working princi⁃ple and simulation results of each module of the software are proposed,and the results show that the method can effectively identify multiple emitters.