AO2DS: A Method of Auxiliary Operational Decision-making Based on System Dynamics Simulation

The intelligent decision plays an important role in the modern war. This paper proposes AO2DS, an auxiliary operational decision method based on system dynamics simulation, to assist commander making decision. The State-of-the-Art system dynamics simulation method relies on complex models and logical relationships. AO2DS reduces the complexity by extracting the essential decision points based on domain knowledge. Meanwhile, AO2DS solves the problem of model explosion by using modular modeling and multi-view segmentation. Furthermore, we expand a dynamic link libraries in Vensim, which help AO2DS to achieve complex calculations and logical judgments. We have further applied AO2DS on The StarCraft II game platform. Our experimental evaluation shows AO2DS can defeat the Difficult computer, with more than 80% winning rate.