Intelligent radiation source identification method based on combined twin network

The invention discloses an intelligent radiation source identification method based on a combined twin network, and the method specifically comprises the steps: S1, collecting a plurality of samples, and carrying out the data enhancement through employing the twin network; s2, according to the enhanced data, adopting a trained combined twin network to identify an unknown target; s3, calibrating an unknown target, and updating the combined twin network according to the calibrated unknown target; and S4, carrying out reentry identification on an unknown target by adopting the updated combined twin network. On the basis of fully fusing fingerprint information of traditional signal identification, hidden features of the radiation source signal are further mined through the deep learning network, and the problem that an existing radiation source identification method is low in accuracy is effectively solved. 本发明公开了一种基于组合孪生网络的智能辐射源识别方法，具体为：S1、采集若干样本，使用孪生网络进行数据增强；S2、根据增强后的数据，采用训练好的组合孪生网络对未知目标进行辨识；S3、对未知目标进行标定，并根据标定后