关键词: 计算机生成兵力, 行为建模, 贝叶斯网络, 深度学习, 强化学习, 知识图谱

Abstract: System simulation is an effective means to solve many military problems under the condition of informationization. In the simulation system, the accuracy of computer generated force behavior is an important factor that affects the conclusion of military research. Considering that the autonomy of CGF is difficult to meet the needs of military research, CGF behavior modeling methods in the past 10 years are systematically summarized, the advantages and disadvantages of different methods are compared and analyzed, the development status of CGF behavior modeling methods at home and abroad in recent five years is summarized. Aiming at the research demand of military issues, the existing problems and development prospects in this field are summarized and prospected, and six CGF behavior modeling thinking methods are proposed.

Key words: computer generated force, behavior modeling, Bayesian networks, deep learning, reinforcement learning, knowledge graph