摘要
以分布式认知任务分析与多模态数据驱动为理论基础,在通用智能导学框架的支持下设计外部评估引擎,综合利用机器学习算法分析处理视频、语音和模拟训练日志等信息,实现了城市反恐作战中室内清剿典型场景任务分队作战效能评估。该研究体系架构健全,智能化程度较高,扩展性较强。通过对该项目体系架构、关键技术和使用效果进行深入研究,得出相关结论,为分队级作战效能评估提供参考。
In recent years,the U.S.Army Combat Capabilities Development Command and the Vanderbilt University have jointly carried out research on how to evaluate the efficiency of urban counter-terrorism detachments in a training environment that combines virtual and real scenarios.The joint research program is aimed to enhance the combat capability of the US Army in urban counter-terrorism operations and ensure its ability of delivery.Based on the theoretical foundation of distributed cognitive task analysis and multi-modal data,an external evaluation engine has been designed with the support of the General Intelligence Framework for Tutoring,which comprehensively utilizes machine learning algorithms to process such information as video,voice and simulated training logs.This helps evaluate the combat efficiency of a typical indoor clearance task in urban counter-terrorism operations.The system has a sound architecture,a high degree of intelligence and goodscalability.An in-depth study of the system architecture,key technologies and effects is conducted and conclusions have been drawn that can provide references for the evaluation of combat efficiency at the detachment level.
作者
叶磊
孙磊刚
王千
胡海
YE Lei;SUN Leigang;WANG Qian;HU Hai(Test and Training Base,National University of Defense Technology,Xi’an 710100,China;College of Intelligence Science and Technology,National University of Defense Technology,Changsha 410073,China;Combat Service Security Brigade,Staff Department of the Armed Police Force,Beijing 100089,China)
出处
《国防科技》
2024年第4期134-142,共9页
National Defense Technology
关键词
城市作战
反恐分队
效能评估
认知任务分析
多模态数据驱动
urban combat
counter-terrorism detachment
efficiency evaluation
cognitive task analysis
multi-modal data-driven
