Research on Realtime Decision Making Strategy of Swarm Intelligence based Gaming

Modern war has the features of multi-level, high dynamic, high coupling and high complexity. intelligent situation cognition and tactical decision making based on hierarchical reinforcement learning are the future development trend. To face the current challenges, this paper studies the lightweight problem of situation cognition and realtime decision making strategy in complex incomplete information dynamic game, focusing on game space reduction, multi-level situation cognition and collaborative strategy optimization. The game decision making model driven by data and knowledge is constructed, and the level of game strategy is improved through self-gaming of reinforcement learning, to gradually reach the level of human commander.