基于空间分块的 CGF 注意力感知方法
Perception Method of CGF Attention Based on Spatial Partitioning Strategy
-
摘要: 陆军分队战术级作战仿真系统中计算机生成兵力(computer generated forces, CGF)的战场行为与地理环境特征和其他实体活动紧密相关,而优化CGF对象的虚拟战场环境感知、对象管理和查询方法, 则是提升其行为模型可信度及仿真系统运行效率的关键所在. 着眼观察- 判断- 决策- 行动(observe-orient-decide-act, OODA)循环中的观察环节及判断环节, 基于空间分块思想实现了对战场空间的高效感知及动态管理;结合多感官感知注意力机制(视、听)及感知主体实例偏好, 应用价值排序建立了基于感知注意力机制的价值模型, 并与传统感知方法进行了横向对比;结合空间分块和价值模型搭建了分队战术级态势感知仿真平台.基于平台仿真结果, 得出结论: 基于空间分块的CGF 注意力感知方法在不损失感知精度的前提下能够有效提高感知效率, 具有可行性和有效性.Abstract: Computer generated forces(CGF)behavior in the combat simulation system for element tactics training is closely related with geographical environment characteristics and activities of other entities. The optimization of virtual battlefield environment perception, object management and inquiry method of CGF entities is a key in enhancing confidence level of behavior model and operational efficiency of simulation system. Focused on the perception observation and judgment link in the observe-orient-decide-act (OODA)cycle, firstly, the efficient perception and dynamic management of battlefield space based on the concept of space block is realized. Then, attention mechanism of multi sensory perception (visual and auditory)and instance preference of perceptual subjects are combined, a value model based on perceived attention mechanism is established with ordering of values, thus the horizontal comparison between traditional perception model and the proposed model is conducted. Finally, a tactical-levelsimulation platform of element situational awareness is built based on space block and value model. Based on the simulation results, the conclusion is that the perception method of CGF attention based on spatial block strategy can effectively improve the perception efficiency without losing perception accuracy, it is feasible and effective.
下载: