摘要
为实现空管员直接发布指令来操控无人机,结合深度学习在自然语言处理(natural language processing,NLP)中的应用,提出一种基于深度学习的无人机指令意图识别方法。使用改进Skip-Gram模型生成指令文本的词向量,输入到卷积神经网络进行指令文本分类,得到空管员发布指令的意图。通过实验验证,结果表明:该方法能够较准确地对指令意图进行识别,有助于后续指令理解技术的实现,为进一步实现空管员与无人机直接交互做准备。
In order to realize that air traffic controllers can directly issue instructions to control unmanned aerial vehicles(UAVs),combined with the application of deep learning in natural language processing(NLP),a method of unmanned aerial vehicle(UAV)command intention recognition based on deep learning is proposed.The improved Skip-Gram model is used to generate the word vector of the instruction text,which is input into the convolutional neural network to classify the instruction text,and the intention of the air traffic controller to issue the instruc tion is obtained.The experimental results show that the method can accurately identify the command intention,which is helpful for the realization of the subsequent command understanding technology and for the further direct interaction between air traffi c controllers and UAVs.
作者
符凯
朱雪耀
吕全喜
姜超
Fu Kai;Zhu Xueyao;LYU Quanxi;Jiang Chao(AVIC Xi’an Flight Automatic Control Research Institute,Xi’an 710065,China;Aviation Key Laboratory of Science and Technology on Aircraft Control,Xi’an 710065,China)
出处
《兵工自动化》
2022年第10期41-44,59,共5页
Ordnance Industry Automation
关键词
无人机
空地对话
自然语言处理
意图识别
卷积神经网络
UAV
air-to-ground dialogue
natural language processing
intent identification
convolutional neural networks