Research on Intelligent Blue Army Strategy Modeling Method Based on Distribution Learning

In the context of intelligent warfare, constructing realistic and professional Blue Forces has become an essential means to enhance the training and combat capabilities of Red Forces. This paper presents an intelligent Blue Army strategy modeling method based on distribution learning, which ap-proximates potential Blue Force behaviors in actual battlefield environments by combining the probability distributions of var-ious strategy types. This method overcomes the limitations of traditional fixed-opponent modeling and enhances the flexibility and diversity of Blue Force strategies through techniques such as reinforcement learning and self-play. Based on practical expectations, this method effectively improves the realism and adversarial nature of intelligent Blue Forces, fulfilling future combat training requirements.