知识和数据协同驱动的智能作战任务规划框架OA
Intelligent Combat Mission Planning Framework Driven by Knowledge-data Synergy
现代战争信息化程度持续增加,传统的任务规划方法已经难以适应智能决策的需求.为提升我军作战设计能力,占据未来智能化战场的主动地位,提出一种知识和数据协同驱动的智能作战任务规划框架,将数据驱动的态势抽象及行动预测方法、知识驱动前向搜索的行动序列生成方法、基于大语言模型的样本增强和网络迭代训练方法融入OODA环的不同阶段,从而实现智能作战任务规划.
With the continuous improvement of informatization level in modern warfare,traditional mission planning methods can hardly meet the requirements of intelligent decision-making.To enhance the combat design capability of our military and seize the initiative in future intelligent battlefields,this paper proposes a knowledge-data synergy-driven framework for intelligent combat mission planning.The framework integrates data-driven situation abstraction and action prediction,knowledge-driven forward search for action sequence generation,as well as large language model-based sample augmentation and network iterative training into different stages of the OODA loop,so as to realize intelligent combat mission planning.
邵天浩;张青;张可;程恺
陆军工程大学指挥控制工程学院,南京 210000陆军工程大学基础部基础实验室,南京 210000国防科技大学第六十三研究所,南京 210000陆军工程大学指挥控制工程学院,南京 210000
信息技术与安全科学
知识和数据协同驱动作战任务规划OODA环层次任务网蒙特卡洛树搜索抽象技术大语言模型
knowledge-data synergy-drivencombat mission planningOODA loophierarchical task networkMonte Carlo tree searchabstraction technologylarge language model
《火力与指挥控制》 2026 (5)
1-12,12
国家自然科学基金资助项目(61806221)
评论