海空兵棋中基于行为树的智能决策方法

Intelligent Decision-Making Method based on Behavior Tree in Sea-Air Wargames

在兵棋推演过程中,面对策略的不确定性和决策情况的复杂性,人类专家制定的策略往往需要经过计算推演的验证,计算结果也可能会和预想结果大相径庭,而智能决策辅助推演可以极大地提升推演过程中策略制定的准确性。针对传统有限状态机决策机制的不灵活,模糊状态机又缺乏适应性的问题,提出了一种基于行为树的智能决策方法来实现兵棋的实时决策过程,并基于Behavior3 Editor平台实现行为树的策略生成过程,最后将基于行为树的智能决策方法用于海空兵棋推演,实现了作战策略的生成,验证了该方法的可行性和有效性。

In the process of wargaming, with the uncertainty of the strategy and the complexity of the decision-making situation,strategies formulated by human experts often need to be verified by computational deduction, and the calculation results may be very different from the expected results, intelligent decision-making can greatly improve the accuracy of strategy formulation during the wargaming process. To overcome the inflexibility of traditional finite state machine decision-making mechanism and the lack of adaptability of fuzzy state machine, an intelligent decision-making method based on behavior tree is proposed to realize the real-time decision-making in wargames, and the strategy of behavior tree is realized based on the platform of Behavior3 Editor. This method is applied to sea-air wargaming to give combat strategy, and the feasibility and effectiveness of the method have been verified.