Abstract: 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.