基于改进知识推理的联合打击目标关联推荐

Joint Strike Target Association Recommendation Based on Improved Knowledge Reasoning

联合作战指挥中,如何针对指挥员的选定打击目标,由指挥系统后台智能推荐符合指挥员倾向意图的关联目标,是亟待解决的实际问题。利用目标实体之间的隶属关系,构造符合战场信息传播原则的多元实体及关系集,引入知识图谱推理的向量优化技术补全目标超网络的隐藏关系,进而构建打击目标超网络体系并实现打击目标的关联推荐。最后通过仿真实验对比分析了改进方法与常规超网络分析方法的异同,探讨了方法的正确性和优越性。方法并未使用专家主观评分法,用知识推理取得了和其他同类方法相似的体系价值评分结果,在智能化的目标关联推荐和作战辅助决策中具有较好的应用前景。

In joint combat command,how to strike the targets according to the commander’s selection and recommend an associated target intelligently by command system background conforming to the commander's intention is a practical problem needed to be solved.The affiliation among target entities is utilized to construct the multi entities and relation sets according to the information dissemination principle of battle field.The vector optimi zation technology of knowledge graph reasoning is introduced to complete the hidden relationship of target hyper network so that the target hyper network system can be constructed and the association recommendation of striking the targets can be realized,finally,the similarities and differences between the proposed method and conventional supernetwork analysis method are compared with the simulation experiments.The correctness and the superiority of the proposed method are explored.The subjective test method of experts is not used in the proposed method,the knowledge reasoning is used to obtain the system value evaluation results similar with other same types of methods,so the proposed method has a better application prospect in the intelligent target association recommendation and assistant decision-making of combat.