基于深度学习的空中任务识别方法研究

发布时间：2018-12-21 14:31

【摘要】：在大规模兵棋仿真推演中,空中任务是指挥员关注的重点。对空中任务的快速、准确和自动识别,是智能决策的前提和基础。深度学习技术的迅速发展,为复杂战场态势特征提取提供了现实可行的解决方法,为研究空中任务识别提供了技术支持。概述了传统任务识别研究方法和基于深度学习的任务识别方法研究进展,分别对卷积神经网络(Convolutional Neural Networks,CNN)、长短时记忆网络(Long Short Term Memory,LSTM)、生成对抗网络(Generate Adversarial Network,GAN)3种深度学习方法在空中任务识别问题中的应用进行了论述,提出了解决思路。
[Abstract]:In the simulation of large-scale military chess, the air task is the focus of the commander's attention. The fast, accurate and automatic recognition of air missions is the premise and foundation of intelligent decision-making. The rapid development of depth learning technology provides a practical solution for complex battlefield situation feature extraction and provides technical support for the research of air mission recognition. The research progress of traditional task recognition method and task recognition method based on deep learning is summarized. The convolutional neural network (Convolutional Neural Networks,CNN) and long and short memory network (Long Short Term Memory,LSTM) are used to generate (Generate Adversarial Network,. The application of three depth learning methods in air task recognition is discussed, and the solutions are put forward.
【作者单位】：国防大学信息作战与指挥训练教研部;
【基金】：国家自然科学基金(U1435218,61403401)
【分类号】：E919;TP18

【相似文献】