基于RBF神经网络的雷达干扰效能评估  

Radar Jamming Effectiveness Evaluation Based on RBF Neural Network

作　　者：袁学华[1] 李龙骧[1] 

机构地区：[1]电子工程学院,合肥230037

出　　处：《电子对抗》2016年第2期14-18,共5页Electronic Warfare

摘　　要：利用RBF神经网络算法对雷达干扰效能进行评估。首先，确定雷达干扰效果评估的指标体系及其隶属度函数，并以各指标的隶属度函数值作为神经网络输入层数据；然后，根据选定的训练样本对RBF神经网络进行训练，再运用测试样本对该神经网络性能进行检验，并与BP神经网络算法进行了比较。理论推导与仿真实验表明了该方法的可行性与有效性，且比BP神经网络具有更快的收敛速度和更好的稳定性。Radar jamming effectiveness evaluation is a complicated network is used for radar jamming effectiveness evaluation, and the jamming effectiveness factors are determined according to the factors＇ problem, the RBF neural degrees of membership for sub-jection functions, so the degrees can be regarded as input data for RBF neural network. Then training samples are chosen to train the RBF neural network, and the testing samples are used to check out the per- formance. Theoretical analysis and computer simulation show that： the new way has faster con- vergence rate and better stability, compared to the BP neural network algorithm.

关 键 词：雷达干扰效能评估 RBF神经网络 BP神经网络 隶属度函数 

分 类 号：TN974[电子电信—信号与信息处理]