Radiation source identification method based on deep learning and ds evidence theory fusion

Abstract

The application relates to a radiation source identification method based on deep learning and DS evidence theory fusion. The method comprises the following steps: collecting and obtaining electromagnetic signals of a to-be-identified radiation source in a target area, and performing pretreatment, signal representation and data enhancement processing, and randomly dividing the signals into a training set, a verification set and a test set according to a certain proportion; inputting the training set and the verification set into a radiation source identification model containing three types of deep learning networks, training an optimal identification model; inputting the test set into the optimal identification model to obtain identification results output by the three types of deep learning networks; and fusing the identification results output by the three types of deep learning networks based on a DS evidence reasoning theory to obtain a final radiation source type identification result. The method reduces the requirements for the number of signal samples and the noise intensity through multi-type network fusion training, and utilizes the DS evidence reasoning to avoid single network identification deviation, so that the practicability and accuracy of radiation source identification are finally improved.

Images