Research on Hybrid Intelligence Wargame Method
Xin Jin, Xinnian Wang, Ran Ding, Yunchao Wu · 2022
Wargame, as a tool to generate sample data for analysis and model training, has vast application in fields of training, command & control, and tactical research. Traditional wargame technologies greatly rely on human wisdom in the loop, impossible to generate large scale sample data. Reinforcement learning technology can generate large scale sample data, but it is not competent for the decision complexity above campaign level. This paper proposes a hybrid intelligence wargame method, which can generate large scale sample data using AI algorithms under the guidance of human wisdom. It has wide applications, which provides data analysis functions that existing wargame methods cannot provide. Prototype software has been developed based on the method, with feasibility and effectiveness verified through experiments, which has certain reference value.