非结构化网格下海洋流场的特征提取与种子点选取算法  

Feature Extraction and Seed Point Selection Algorithms of Ocean Flow Field on Unstructured Grids

作　　者：李忠伟[1] 宫凯旋 李永[1] 刘格格 LI Zhongwei;GONG Kaixuan;LI Yong;LIU Gege(College of Oceanography and Space Informatics,China University of Petroleum(East China),Qingdao 266580,Shandong,China;College of Computer Science and Technology,China University of Petroleum(East China),Qingdao 266580,Shandong,China)

机构地区：[1]中国石油大学(华东)海洋与空间信息学院,山东青岛266580 [2]中国石油大学(华东)计算机科学与技术学院,山东青岛266580

出　　处：《计算机工程》2024年第1期320-328,共9页Computer Engineering

基　　金：国家自然科学基金重点项目(62231028);国家重点研发计划(2018YFC1406204);中央高校基本科研业务费专项资金(22CX01004A-9)。

摘　　要：对于非结构化网格流场数据,现有研究存在不能定位流场中临界点的具体位置、无法对临界点精细分类以及缺乏适用于非结构化网格的种子点选取算法等问题。针对上述问题,基于非结构化网格流场数据,分别在临界点提取和种子点选取方法上进行改进创新。提出非结构化网格中临界点的定位及分类方法,通过庞加莱指数法判断存在临界点的三角网格。构造质心迭代法定位临界点在网格中的准确位置,并设计三角网格内雅克比矩阵的构造方法,将临界点精细分类。基于非结构化网格提出基于最大得分和网格密度的种子点选取算法,先比较相邻格点的标量值大小来计算每个格点的得分,形成“最大得分”标量场,再按照网格密度动态设置阈值,将得分大于阈值的格点选为种子点,接着以种子点为起始点生成流线,生成的流线可以表达出流场的关键特征与全局信息。基于多个海域流场数据的实验结果表明,临界点分类的准确率可达99%以上,证明了临界点提取算法的准确性以及种子点选取算法对提升流场可视化效果的有效性。In the context of unstructured grid flow field data,existing research faces challenges such as the inability to precisely locate critical points within the flow field,a lack of fine classification for these points,and insufficient methods for seed point placement on unstructured grids.To address these issues,this study introduces innovative improvements in the extraction of critical points and selection of seed points,specifically tailored to unstructured grid flow field data.Initially,an innovative technique for locating and classifying critical points on unstructured grids is proposed.The Poincare index method is employed to identify triangular grids containing critical points.Following this,a center-of-mass iteration method is developed to accurately locate critical points within these grids.Additionally,a method for constructing a Jacobian matrix in triangular grids is designed for the fine classification of critical points.Subsequently,the study introduces a seed point selection algorithm for unstructured grids,based on a"maximum score"principle and grid density.This involves calculating a score for each grid point by comparing the scalar values of adjacent points,thereby creating a"maximum score"scalar field.A threshold value is then set based on grid density,and grid points exceeding this threshold are selected as seed points.These seed points serve as the starting points for generating streamlines,effectively capturing key features and conveying the global information of the flow field.Finally,a comparative analysis of experimental results from flow field data across several seas is presented.This analysis demonstrates that the accuracy of critical point classification can exceed 99%,validating the precision of the proposed critical point extraction method and the effectiveness of the seed point selection algorithm in enhancing the visualization of flow fields.

关 键 词：非结构化三角网格 临界点提取 种子点选取 最大得分 流场可视化 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]