面向联合作战兵棋推演的智能决策建模框架

An Intelligent Decision-making Modeling Architecture for Joint Operation-oriented Wargaming

围绕联合作战兵棋推演的智能化需求,提出一种战役战术分层结构的智能决策建模框架.模拟作战控制过程,框架通过作战行动序列引导约束决策行为生成,并基于任务式指挥思想实现战役层对战术任务的控制调度.重点研究了战役层智能决策模型的输入输出定义、策略生成过程和学习进化方法,探讨了面向兵棋推演的作战行动序列表示、目标驱动的自适应作战控制建模、面向博弈策略优化的任务效果预测等关键技术的实现途径.所提框架具有一定的可扩展性和实用性,能够在基于兵棋推演的军事应用场景中发挥价值.

Aiming at the intelligent requirements for AI-enabled joint operational wargaming,an intelligent decision-making modeling architecture with a campaign-tactical hierarchical structure is proposed.The combat command decision-making process is simulated,the architecture is guided by the courses of action to constraint the generation of the decision-making behaviors,the control and scheduling of tactical tasks on the campaign level is realized based on the mission command concepts.and schedules tactical tasks through mission commands.Its input and output definitions,the strategy generation process and learning evolutionary methods are studied in details the approaches of the realization of such key technologies as the formal courses of action representation for wargaming,goal-driven adaptive combat control modeling,and task effect prediction for gaming strategy-oriented optimization,etc.are explored.The presented architecture addresses certain level of extensibility and practicality,and can play its role for military AI-enabled wargaming application scenes.