基于时序图的作战指挥行为知识表示学习方法

为深入探索时序作战指挥行为知识的建模方法,实现对具有时序关联特征的指挥员作战指挥行为的有效表征,以兵棋推演作战指令为基础,提出一种基于时序图的作战指挥行为知识表示学习方法,对作战指挥行为进行知识表示学习,并通过作战指挥行为预测任务验证模型的有效性。实验结果表明,提出的方法对于评估指标提升较大,能够有效捕捉想定场景下联合作战指挥员的作战指挥行为时空特征,为时序作战指挥行为知识的表示学习提供了可行范例,为联合作战指挥员的指挥经验抽取和联合作战态势认知提供基础。

Operations command behavior knowledge representation learning method based on sequential graph

In order to explore the modeling method of the sequential operations command behaviors knowledge deeply and effectively obtain the temporal correlation features of operations command behaviors of the commanders, an operations command behaviors knowledge representation learning method is proposed based on the sequential graph of wargaming operational orders. The operations command behaviors are presented by knowledge representation learning, and the model is verified by the task of operations command behaviors prediction. Experimental results show that the evaluating index is increased obviously, and the spatial-temporal features of operations command behaviors by the joint operations commanders are obtained effectively. Our work highlights the sequential joint operations command behavior representation learning and provides a foundation for the command experience extraction of joint operations commanders and joint operations situation cognition.