基于灰色关联度神经网络的雷达型号识别模型

针对现代电子战对雷达目标信号的复杂性和残缺性以及实用雷达目标识别系统的健壮性和扩展性等要求,提出一种基于灰色关联度和BP神经网络的灰色神经网络识别模型.首先采用比较成熟的BP神经网络对侦察雷达目标信号进行粗分,识别出雷达的体制;然后把模板数据库中该体制的雷达标准数据作为比较序列,建立差异信息空间,再把观测的数据和比较序列进行灰关联度分析,得出其对应的关联度,从而识别出雷达的具体型号.仿真结果表明在对参数残缺或畸变以及新体制的雷达辐射源进行识别时,取得良好的效果.表明综合灰色神经网络对辐射源进行识别是完全可行的,并且可以提高识别率、可靠性.

Model of Radar Type Identification Based on BP Neural Network Grey Relation Theory

According to the complexity and incompleteness of radar target signals and robustness,extensibility requirements of practical radar target recognition systems, a grey neutral network recognition model based on BP neural network is proposed.The first part of this model,BP neural network,which has been well studied,is used to recognize the radar system;then a comparable sequence is established according to the standard radar data of the recognized radar system in radar modal database.Grey relation degree of the radar data and comparable sequence can be attained.By the grey relation degree,specified radar type can be identified.The simulation results show that this model is effective for identifying incomplete parameters and new radar types.It proved that the composition of grey neutral network used to identify radar target can increase the identification rate and robustness.