【摘要】 以战场指挥决策智能化为目标，针对无人系统集群作战对抗过程，研究战场态势理解问题。构建层次型态势理解框架，按照个体和群组两个层次分层处理动态战场信息。针对个体威胁评估过程具有多源输入特点的问题，采用极限学习机方法实现对敌方个体威胁评估的推理；针对群组层次需要对敌方作战单元进行合理分组的问题，采用K均值聚类算法得到群组态势分析，实现战场态势的深入挖掘，从而更有效地处理复杂的战场信息。更多还原

【Abstract】 Aiming at the intelligence of battlefield command and decision-making,the battlefield situation understanding problem in the process of cluster combat confrontation of unmanned system is studied. A hierarchical situation understanding framework is constructed,which processes dynamic battlefield information hierarchically according to individual and group levels. According to the characteristics of multi-source input in the process of individual threat assessment,extreme learning machine method is used to realize the reasoning of enemy individual threat assessment. As the group hierarchy needs the enemy’s combat unit to make K-means clustering algorithm is used to get the group situation analysis to make reasonable grouping,so as to realize the in-depth mining of battlefield situation and to deal with complex battlefield information more effectively.更多还原