一种基于贝叶斯网络的雷达重频模式识别方法

雷达重频模式指雷达脉冲重复间隔（PRI）的调制样式，重频模式识别对于雷达型号、类别识别具有重要支持作用，但这是一个较困难的过程，一般难以通过单一重频特征完成多种复杂重频模式的自动识别。本文为解决雷达多重频模式自动识别及识别抗噪声干扰问题，针对多种重频模式提取了几种PRI序列特征量（重频特征），然后引入了贝叶斯多网络分类器（BMNClassifier），利用所提取的特征量作为贝叶斯多网络分类器输入节点，通过贝叶斯多网络分类器的概率推理能力来实现雷达重频模式识别。

Approach to Radar Pulse Repetition Intervals Modulation Recognition Based on Bayesian Networks

The pattern of radar pulse repetition intervals(PRI) is also called radar pulse repetition intervals modulation.The recognition of radar pulse repetition intervals modulation can make a great support to both recognition of radar's types and categories.But it is a difficult process.Generally the multi-pattern of radar pulse repetition intervals cannot be recognized automatically only via one character of PRI.In order to recognize PRI's patterns automatically and to reduce the influence of data noise,some PRI's characters,which just aim at these PRI's patterns,are extracted.Then the Bayesian multi-net(BMN) classifier is introduced and the PRI's characters are used as its input nodes.So the PRI's patterns are deduced by conditionally probability via BMN classifier.