考虑多维关联性的结构化数据文件三维可视化分析研究  

Research on 3D Visualization Analysis of Structured Data Files Considering Multidimensional Correlation

作　　者：马永 赵煜阳 雷霆 张靖 王俊 Ma Yong;Zhao Yuyang;Lei Ting;Zhang Jing;Wang Jun(State Grid Anhui Electric Power Co.,Ltd.Information and Communication Branch,Hefei,Anhui 230061)

机构地区：[1]国网安徽省电力有限公司信息通信分公司,安徽合肥2300610

出　　处：《现代科学仪器》2024年第5期297-303,共7页Modern Scientific Instruments

基　　金：国网安徽省电力有限公司科技项目(521207220002)。

摘　　要：数据格式间的浮点位数差异的忽略会导致三维可视化出现误差。为此,提出考虑多维关联性的结构化数据文件三维可视化研究。利用映射原理对源数据重构,引入增益矩阵消除数据格式间的浮点位数差异,转换数据文件头格式。求取数据多维关联度,构建三维空间数据体,对数据网格化处理,利用SynGlyphX和Tableau数据可视化数据库输出。实验显示,所提方法在不同数据量条件应用下,每秒绘制帧数均在50FPS以上,且三维可视化结果准确性更高,效果更优。由此说明,所提方法在考虑数据之间的多维关联性的同时,能够更准确地捕捉和表达数据之间的复杂关系。Ignoring floating-point bit differences between data formats can lead to errors in 3D visualization.Therefore,a three-dimensional visualization of structured data files considering multi-dimensional correlation is proposed.By using the mapping principle,the source data is reconstructed,the gain matrix is introduced to eliminate the difference of floating point bits between data formats,and the format of data file header is converted.The multi-dimensional relational degree of data is obtained,the three-dimensional spatial data volume is constructed,the data is processed by grid,and the data is output by SynGlyphX and Tableau data visualization database.Experiments show that the proposed method can draw more than 50FPS per second under different data volume conditions,and the 3D visualization results are more accurate and the effect is better.Therefore,the proposed method can capture and express the complex relationship between data more accurately while considering the multi-dimensional correlation between data.

关 键 词：多维关联性 结构化数据文件 三维可视化分析 数据三维交互 

分 类 号：TD73[矿业工程—矿井通风与安全]