舰船电子工程2019，Vol.39Issue(5)：26-29,4.DOI:10.3969/j.issn.1672-9730.2019.05.008
一种TensorFlow平台上目标意图识别模型设计与实现∗
A Target Intent Recognition Model Designed and Implemented on TensorFlow Platform
摘要
Abstract
According to the attribute characteristics of the target recognizes the target intent,which is very important for the battlefield to make the next decision. The traditional method of intent recognition requires a lot of prior knowledge and it is difficult to deal with timing problems. The paper adopts a new method of target intent recognition. The method simulates the attribute data of the target at a continuous moment,and performs data preprocessing,and uses the LSTM neural network model for training. Through the simulation data,build and train the LSTM neural network model on the TensorFlow platform,and finally determine the structure and parameters of the network. Experimental result shows that the constructed LSTM network model achieves better results.关键词
意图识别/模型设计/时序问题/LSTM网络/数据处理Key words
intention recognition/model design/sequence problem/LSTM network/data processing分类
信息技术与安全科学引用本文复制引用
李宁安,张剑,周倜..一种TensorFlow平台上目标意图识别模型设计与实现∗[J].舰船电子工程,2019,39(5):26-29,4.基金项目
装备发展部预研课题项目(编号:315080404) （编号:315080404）
海军装备部预研课题项目(编号:302010102105)资助. （编号:302010102105）
