Find in Library

With the development of science and technology and the change of operational methods, the reasoning of the enemy’s operational intent has begun to be introduced into the battlefield. Air defense operational decision-making has higher requirements for the accuracy of intent identification. The existing knowledge atlas, expert networks, and deep neural network methods still have gaps in the accuracy of identification, which is difficult to meet the requirements of air defense operations. Therefore, this paper combines the advantages of convolutional neural network and cyclic neural network to design a deep learning model FCN-BiGRU for air target intention recognition. The Full Convolutional Network (FCN) can extract the complex local features in air combat data, and the Bidirectional Gated Recurrent Unit (BiGRU) is used to capture the temporal characteristics of air combat intention data. It is proved by ablation experiments and comparative experiments that the accuracy rate of intention recognition of the FCN-BiGRU model far reaching 98.71% and 1.14% higher than the existing air target intention recognition model, which provides a more powerful basis for air defense combat decision-making.