基于时空要素的丢失战场态势海空目标还原推荐应用  

Lost Battlefield Sea And Air Target Reconstructive Application Recommendation Based on Temporal-Spatial-Element

作　　者：陈娜 胡佳浩 Chen Na;Hu Jiahao(The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210007,China)

机构地区：[1]中国电子科技集团公司第28研究所,南京210007

出　　处：《信息化研究》2025年第2期43-49,56,共8页INFORMATIZATION RESEARCH

摘　　要：本文针对战场态势丢失海空目标的还原推荐,引入时空要素编码技术,对目标轨迹进行网格化构建,实现多维数据到一维的时空要素网格编码集合的转换,提出了一种基于时空网格模型的数据索引方法,实现数据特征的统一表示与组织。采用长短时记忆网络算法,对丢失的战场态势海空目标动态预测,结合分级剖分索引查询方法,快速形成可疑目标清单,辅助指挥员及时判断,作出处置。对开源数据开展实验验证,验证了基于时空要素编码的目标索引构建方法在海量数据情况下,具有时间优势,为数据组织提供一种可行的方法。This article addresses the issue of recovering lost battlefield sea and air targets by introducing spatiotemporal feature encoding technology to construct a gridded representation of target trajectories.This approach achieves the transformation of multidimensional data into a one-dimensional collection of spatiotemporal feature grid encodings.A data indexing method based on a spatiotemporal grid model is proposed,enabling the unified representation and organization of data features.Using a Long Short-Term Memory(LSTM)network algorithm,dynamic prediction of lost battlefield sea and air targets is made.Combined with a hierarchical partitioning indexing query method,a list of suspicious targets can be quickly generated,assisting commanders in timely assessments and decision-making.Experimental verification of open-source data is carried out to verify that the target index construction method based on temporal-spatial-element has a time advantage in handling massive datasets,providing a feasible approach for data organization.

关 键 词：时空要素编码 轨迹网格化 剖分索引 时空关联 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]