现代战争的军事作战数据来源广泛多样,目前对于军事作战的数据模型研究主要针对其时空属性进行描述,对于作战实体的多样化特征缺乏完整的表达,难以挖掘多维度的作战信息。多粒度时空对象建模是依照多粒度时空对象数据模型框架,将时空实体简化和抽象为数据模型,对复杂的现实世界进行概括与表达。本文在分析作战实体信息和特点的基础上,依托多粒度时空对象建模的思路,建立作战实体对象类并构建作战对象模型,对战争中时空实体的关联关系、组成结构和行为能力等特征的抽象与表达进行了探索,全面多维地描述作战实体及其特征变化,并以阿富汗战争中的美军凯克提特遣队为例进行对象化建模与可视化实现。实验结果表明,多粒度时空对象建模方法能有效表达作战实体的多维特征,在战场信息挖掘上具有精确立体、全面多维的优势,为面向对象的精细化作战研究提供了理论基础。

At present, military combat data for modern warfare come from a wide variety of sources. However, the research on data modelling of military operations mainly focuses on the description of its spatio-temporal attributes. It lacks complete description for the diversified characteristics of combat entities and is difficult to mine multi-dimensional combat information. According to the framework of multi-granularity spatio-temporal object data model, spatio-temporal entities are simplified and abstracted as data models to summarize and express the complex real world. On the basis of analyzing the information and characteristics of combat entities, this paper builds the combat entity object classes and object models based on the idea of multi-granularity spatio-temporal object modeling, and carries out the preliminary study of the abstraction and expression of the characteristics of spatio-temporal entities, such as association relationship, composition structure, and behavior. We take the U.S. Kaiketi Task Force in the Afghan war as an example. The experimental results show that the multi-granularity spatio-temporal object modeling method provides a new idea for the expression of multi-dimensional characteristics of combat entities, and can mine the combat information precisely and comprehensively, which provides a theoretical basis for fine combat research.