基于深度学习的联合作战态势智能辅助分析研究

项目名称： 基于深度学习的联合作战态势智能辅助分析研究

项目编号： No.61773399

项目类型： 面上项目

立项/批准年度： 2018

项目学科： 其他

项目作者： 廖鹰

作者单位： 中国人民解放军战略支援部队信息工程大学

项目金额： 16万元

中文摘要： 在信息化战争中，战场空间空前拓展、信息海量化、环境复杂化、作战对象多元化、作战强度急剧增大给指挥人员的精准指挥带来巨大挑战。传统基于人经验或认知的指挥模式已无法有效应对瞬息万变的战场和海量数据，准确快速地把握联合作战的战场态势成为了难题。同时，现有面向作战指挥的计算机辅助决策尚不能以人的认知模式来理解态势，难以适应新形势的需要，人工智能的研究水平已成为制约战场态势理解的瓶颈。复杂网络、大数据和深度学习研究的逐步兴起，为战场态势的智能辅助提供了契机。因此，面对全维战场，研究基于深度学习的联合作战智能辅助分析具有重要意义。本项目从体系对抗性和战争复杂性角度出发，系统研究态势问题，通过参照指挥人员理解态势的模式，集成机器学习方法，基于联合作战条件下的兵棋大数据，建立复合架构的深度学习网络，构建适应性强的态势智能辅助分析模型，实现态势的无监督理解，符合现实的紧迫需求，具有较强的理论和应用价值。

中文关键词： 深度学习；态势；联合作战；兵棋；人工智能

英文摘要： The unprecedented expansion of battlefield space in the information-based war, vast information of battlefield information, complicated battle field environment, diversity of operational object, and large accretion of combat strength poses grave challenges for the commanders. The traditional mode based on experience or cognitive command has been unable to effectively cope with rapidly changing battlefield and huge amounts of data, so to seize the battlefield situation of combined operation accurately and rapidly becomes a difficult problem. At the same time, the existing computer assisting decision-making can’t understand the situation like person and meet the requirements of new situation, so the research of artificial intelligence has become the bottleneck of comprehension of battlefield situation. The rise of research of complex networks，big data and deep learning，provides an opportunity for intelligent assistants of battlefield. Therefore, in the face of the whole omni-dimensional battlefield, research on the intelligence analysis of joint operations based on deep learning is of great significance. This project focuses on combat situation to research the battle situation systematically based on confrontation and complexity system. By reference to the situational model of the commanders personnel understand, integrated machine learning approach, we establish the composite structure of deep learning network based on war game data in the joint operations, to put forward better adaptability intelligent aided analysis model and realize the battle situation unsupervised understanding. This project conforms to the reality of urgent demand, possessing with strong theoretical and application value.

英文关键词： Deep learning ;Posture;Joint Operations ;War games;Artificial intelligence