基于神经网络的辐射源目标快速识别

Radar Radiating-source Recognizing Based on Neural Networks

随着军事科技的快速发展,伪装、隐蔽、欺骗、干扰等手段和技术大量应用于现代战场,以致从传感器获取的目标特征参数的差异越来越不明显,原有的目标识别规律性被大量破坏。在目标特征逐渐模糊的发展趋势下,传统的"特征提取和选择、模板建库、分类器设计、匹配决策"等目标识别技术将很难准确地进行目标识别。论文研究了雷达辐射源识别的基于神经网络快速实现方法,并与传统模板匹配方法进行了效果比对,为摆脱传统目标识别思路的约束,发展智能化的目标识别技术提供了基础。

With the rapid development of military technology,techniques such as disguise,hidden,deception,interference,are applied to the modern battlefield,so the differences in the target feature parameters obtained from the sensor are getting smallerand smaller,the original target recognition regularity is heavily damaged. In the trend of the target to be blurred,traditional“tem？plate match”recognition techniques will be difficult to target accurately. This paper studies radar radiation-source recognizingbased on neural networks,compared with the“template match”,it provides the basis of getting rid of traditional identifying ideasand the development of intelligent target recognition technology.