摘要
为提高雷达辐射源识别系统的识别率,分析了BP神经网络、径向基神经网络和径向基概率神经网络等3种神经网络的结构和性能.用假设的10部雷达参数产生数据进行实验.仿真结果表明,应用径向基概率神经网络能大幅提高雷达辐射源识别的识别率,该网络在雷达辐射源识别中的分类性能明显优于其他2种神经网络。
In order to improve the radar emitter recognition rate, the structure and performance of the BPNN, RBFNN and RBPNN were analyzed. Fulfilled by experiments on the assumed parameters and data out often radars, the results of simulation show that using the RBPNN is to improve greatly the radar emitter recognition rate, and this network is obviously superior to the other two networks in the classification performance on radar emitter recognition.
出处
《空军雷达学院学报》
2007年第1期25-27,共3页
Journal of Air Force Radar Academy
关键词
雷达辐射源
模式识别
神经网络
radar emitter
pattern recognition
neural networks
作者简介
唐健仁（1979-），男，硕士生,主要从事电子对抗信息处理研究．
