Abstract: CGF system is an indispensable technical support of the future fight emulation. Based on introducing the CGF, put forward the design scheme combined higher level decision controllability with underlying mobility autonomy, study the emulation of the anti-aircraft gun element CGF entity in flexible performance, through the study of physical behavior and the intelligent behavior, using BP neural network to realize simulation of intelligent dynamic behavior for the CGF entity, set up the evading module of dynamic obstacle, which solve the problem of colliding of the entity in the course entered of every garage, at last verify the feasibility of the model in consideration of practical application.

Key words: Computer Generated Forces（CGF）, simulation of intelligent dynamic, BP neural network, evading module of dynamic obstacle

摘要： CGF系统是未来作战仿真所必备的技术支持，在介绍CGF系统的基础上，提出了以智能仿真为依托，兼顾底层机动自主和高层决策可控的设计思路。研究了自行高炮连CGF实体在机动性能方面的仿真，通过对高炮机动性能中物理行为和认知行为的研究，运用BP神经网络实现CGF实体智能机动行为的建模，并建立了动态障碍的规避模块。较好解决了中各车行进中出现的实体碰撞问题，结合实际应用对模型进行了可行性验证。

关键词: 计算机生成兵力, 智能机动仿真, BP神经网络, 动态障碍规避