Research on adaptive step size algorithm of marine streamline controlled by information entropy
-
摘要: 海洋流场的流线构造与放置对于认识和理解海洋流场具有重要意义。在流线绘制过程中，积分步长的选择至关重要，能够直接影响流线放置效果。固定步长算法由于无法适应曲率不断变化的情形往往不被采用，以往自适应步长流线算法存在自由度低以及多尺度适用性差的问题。针对上述问题，本文首次将信息熵引入到步长计算中，提出了信息熵控制的海洋流线自适应步长算法。该算法首先通过计算流场信息熵得到熵场，然后依据熵值大小将流场划分为高熵区域和低熵区域，并为每个积分点赋予新的步长，使得流场可以根据变化剧烈程度自适应调整步长大小，即高熵区域（变化剧烈的区域）步长较小，低熵区域（变化平缓的区域）步长较大。实验结果表明，本文算法能够显著增加变化剧烈区域的积分点数目和流线条数，更好地绘制特征处的流线细节，同时在不会影响放置效果的前提下减少了非重要区域的积分点数目和流线条数以提高计算效率，相比以往自适应步长算法显著提高了步长调节的自由度以及尺度适用性，可以应用于不同尺度的海洋流场。Abstract: Abstrart: The streamline construction and placement of the marine flow field is of great significance for recognizing and understanding the marine flow field. In the process of streamline drawing, the selection of integration step is very important, which can directly affect the effect of streamline placement. The fixed step size algorithm is often not used because it cannot adapt to the changing curvature. The previous adaptive step size streamline algorithm has the problems of low degree of freedom and poor multi-scale applicability. In view of the above problems, this paper introduces information entropy into the step size calculation for the first time, and proposes an adaptive step size algorithm of marine streamline controlled by information entropy. Firstly, the entropy field is obtained by calculating the information entropy of the flow field, and then the flow field is divided into high entropy region and low entropy region according to the entropy value, and each integration point is given a new step size, so that the flow field can adaptively adjust the step size according to the intensity of change, that is, the step size of the high entropy region (the region with sharp change) is smaller, and the step size of the low entropy region (the region with gentle change) is larger. The experimental results show that the proposed algorithm can significantly increase the number of integration points and streamlines in the rapidly changing region, better draw the details of the streamline at the feature, and reduce the number of integration points and streamlines in the unimportant region without affecting the placement effect to improve the computational efficiency. Compared with the previous adaptive step size algorithm, the proposed algorithm significantly improves the degree of freedom of step size adjustment and the scale applicability, and can be applied to different scales of marine flow field.
-
图 1 不同种子点放置效果
a.基于特征的种子点放置；b.随机种子点放置；c.均匀种子点放置
Fig. 1 Placement effect of different seed points
a. Places seed points based on features; b. Randomly places seed points; c. Evenly places seed points
图 3 二维流场图及二维熵场图
a.b.c. 二维流场图；d.e.f. 二维熵场图
Fig. 3 Two-dimensional flow field map and two-dimensional entropy field map
a.b.c. Two-dimensional flow field map; d.e.f. two-dimensional entropy field map
图 4 不同固定步长下流线放置效果
a.步长为0.2的固定步长算法；b.步长为0.3的固定步长算法
Fig. 4 Streamline placement effects with different fixed step sizes
a. fixed-step algorithm with a step size of 0.2; b. fixed-step algorithm with a step size of 0.3
图 6 流场A流线放置效果对比
a.步长为0.1的固定步长算法；b.步长为0.2的固定步长算法；c.步长为0.3的固定步长算法；d.考虑速度方向变化的自适应步长算法；e.改进后的AMFCA算法算法；f.本文算法
Fig. 6 Comparison of flow field A streamline placement results
a. fixed-step algorithm with a step size of 0.1; b. fixed-step algorithm with a step size of 0.2; c. fixed-step algorithm with a step size of 0.3; d. adaptive step size algorithm that takes into account changes in velocity direction; e. improved AMFCA algorithm; f. this paper's algorithm
图 7 流场A流线图、温度图和海表面高度图
a.流线图；b.温度图；c.海表面高度图
Fig. 7 Streamline map, temperature map and sea surface height map of flow field A
a. Streamline map; b. Temperature map; c. Sea surface height map
图 8 流场B流线放置效果对比
a.步长为0.1的固定步长算法；b.步长为0.2的固定步长算法；c.步长为0.3的固定步长算法；d.考虑速度方向变化的自适应步长算法；e.改进后的AMFCA算法算法；f.本文算法
Fig. 8 Comparison of flow field B streamline placement results
a. fixed-step algorithm with a step size of 0.1; b. fixed-step algorithm with a step size of 0.2; c. fixed-step algorithm with a step size of 0.3; d. adaptive step size algorithm that takes into account changes in velocity direction; e. improved AMFCA algorithm; f. this paper's algorithm
图 9 流场B流线图、温度图和海表面高度图
a.流线图；b.温度图；c.海表面高度图
Fig. 9 Streamline map, temperature map and sea surface height map of flow field B
a. Streamline map; b. Temperature map; c. Sea surface height map
图 10 流线图相似性对比
a.本文实验结果（流场A）；b.卫星数据流线图（流场A）；c.本文实验结果（流场B）；d.卫星数据流线图（流场B）
Fig. 10 Similarity comparison of streamline map
a. Experimental results in this paper(flow field A); b. Streamline map of satellite data(flow field A); c. Experimental results in this paper(flow field B); d. Streamline map of satellite data(flow field B)
图 11 步长空间分布图
a.流场A部分区域步长空间分布图；b.流场B部分区域步长空间分布图
Fig. 11 Step size spatial distribution map
a. Step size spatial distribution map of flow field A; b. Step size spatial distribution map of flow field B
图 12 三种自适应步长算法对比结果
a.本文算法；b.考虑速度方向变化的自适应步长算法；c.改进后的AMFCA算法
Fig. 12 Comparison of three adaptive step size algorithms
a. this paper's algorithm; b. adaptive step size algorithm that takes into account changes in velocity direction; c. improved AMFCA algorithm
表 1 流场A区域一中不同算法对比结果
Tab. 1 Comparison results of different algorithms in region 1 of flow field A
下载: 导出CSV
表 2 流场A区域二中不同算法对比结果
Tab. 2 Comparison results of different algorithms in region 2 of flow field A
下载: 导出CSV
表 3 流场B区域一中不同算法对比结果
Tab. 3 Comparison results of different algorithms in region 1 of flow field B
下载: 导出CSV
表 4 流场B区域二中不同算法对比结果
Tab. 4 Comparison results of different algorithms in region 2 of flow field B
下载: 导出CSV
表 5 流场A不同算法结果运行时间
Tab. 5 Running time of different algorithm results of flow field A
下载: 导出CSV
-
[1] 王盛波, 潘志庚. 二维流场可视化方法对比分析及综述[J]. 系统仿真学报, 2014, 26(9): 1875−1881,1888.Wang Shengbo, Pan Zhigeng. Comparison, analysis and review of 2D flow visualization[J]. Journal of System Simulation, 2014, 26(9): 1875−1881,1888. [2] 邵绪强, 程雅, 金佚钟. 表意性方法在三维流线可视化中的应用综述[J]. 图学学报, 2022, 43(5): 753−764.Shao Xuqiang, Cheng Ya, Jin Yizhong. A review of the application of illustrative methods in 3D streamline visualization[J]. Journal of Graphics, 2022, 43(5): 753−764. [3] 黄智濒, 傅广涛, 曹凌婧, 等. 基于多视图聚类算法的三维流场关键点附近的流线筛选[J]. 计算机辅助设计与图形学学报, 2022, 34(12): 1930−1942.Huang Zhibin, Fu Guangtao, Cao Lingjing, et al. Streamline selection around critical points of 3D flow fields by the multi-view clustering[J]. Journal of Computer-Aided Design & Computer Graphics, 2022, 34(12): 1930−1942. [4] Liu Fan, Zhou Wensheng, LIU Bingxuan, et al. Flow field description and simplification based on principal component analysis downscaling and clustering algorithms[J]. Frontiers in Earth Science, 2022, 9: 804617. doi: 10.3389/feart.2021.804617 [5] 杨光, 成诗明. 基于四面体的三维流线构造[J]. 北京航空航天大学学报, 2008, 34(9): 1061−1064.Yang Guang, Cheng Shiming. Tetrahedron-based constructing 3D_streamlines in visualization[J]. Journal of Beijing University of Aeronautics and Astronautics, 2008, 34(9): 1061−1064. [6] 鲁大营, 朱登明, 王兆其. 三维流场的流线提取算法[J]. 计算机辅助设计与图形学学报, 2013, 25(5): 666−673.Lu Daying, Zhu Dengming, Wang Zhaoqi. Streamline selection algorithm for three-dimensional flow fields[J]. Journal of Computer-Aided Design & Computer Graphics, 2013, 25(5): 666−673. [7] 季民, 陈丽, 靳奉祥, 等. 自适应步长的海洋流线构造算法[J]. 武汉大学学报•信息科学版, 2014, 39(9): 1052−1056.Ji Min, Chen Li, Jin Fengxiang, et al. Adaptive-step based marine fluid flow streamline constructing algorithm[J]. Geomatics and Information Science of Wuhan University, 2014, 39(9): 1052−1056. [8] 李婷, 季民, 靳奉祥, 等. 海洋流线积分自适应步长计算模型研究[J]. 海洋学报, 2018, 40(3): 95−101.Li Ting, Ji Min, Jin Fengxiang, et al. Research on adaptive-step calculation model of marine fluid flow numerical integration[J]. Haiyang Xuebao, 2018, 40(3): 95−101. [9] Xu Lijie, Lee T Y, Shen Hanwei. An information-theoretic framework for flow visualization[J]. IEEE Transactions on Visualization and Computer Graphics, 2010, 16(6): 1216−1224. doi: 10.1109/TVCG.2010.131 [10] 黄冬梅, 杜艳玲, 张律文. 基于信息熵种子点选取的流线可视化[J]. 计算机工程与科学, 2018, 40(3): 411−417.Huang Dongmei, Du Yanling, Zhang Lvwen. Two information entropy-based seeding methods for 3D flow visualization[J]. Computer Engineering and Science, 2018, 40(3): 411−417. [11] 刘晓帆. 基于信息熵与Clifford代数的流场特征检测[D]. 北京: 北京理工大学, 2016.Liu Xiaofan. Flow feature detection based on entropy and Clifford algebra[D]. Beijing: Beijing Institute of Technology, 2016. [12] 李梦依, 方霞, 郑红波, 等. 基于信息熵的流场定向线积分卷积算法[J]. 计算机应用, 2023, 43(4): 1233−1239.Li Mengyi, Fang Xia, Zheng Hongbo, et al. Oriented line integral convolution algorithm for flow field based on information entropy[J]. Journal of Computer Applications, 2023, 43(4): 1233−1239. [13] 张倩, 肖丽. 基于流线的流场可视化绘制方法综述[J]. 计算机科学, 2021, 48(12): 1−7. doi: 10.11896/jsjkx.201200108Zhang Qian, Xiao Li. Review of visualization drawing methods of flow field based on streamlines[J]. Computer Science, 2021, 48(12): 1−7. doi: 10.11896/jsjkx.201200108 [14] Jobard B, Lefer W. Creating evenly-spaced streamlines of arbitrary density[C]//Proceedings of the Visualization in Scientific Computing’97. Vienna: Springer, 1997: 43-55. [15] Mebarki A, Alliez P, Devillers O. Farthest point seeding for efficient placement of streamlines[C]//VIS 05. IEEE Visualization, 2005. Minneapolis: IEEE, 2005: 479-486. [16] Verma V, Kao D, Pang A. A flow-guided streamline seeding strategy[C]//Proceedings Visualization 2000. VIS 2000 (Cat. No. 00CH37145). Salt Lake City: IEEE, 2000: 163-170. [17] 李忠伟, 徐斌, 李永, 等. 基于非结构化三角网格的海洋流场可视化[J]. 图学学报, 2022, 43(3): 486−495.Li Zhongwei, Xu Bin, Li Yong, et al. Visualization of ocean flow field based on unstructured triangular mesh[J]. Journal of Graphics, 2022, 43(3): 486−495. [18] 牛婵. 基于信息熵的流场特征提取与可视化研究[D]. 秦皇岛: 燕山大学, 2018.Niu Chan. Research on feature extraction of flow field based on information entropy and visualization[D]. Qinhuangdao: Yanshan University, 2018. [19] Du Xiaofu, Liu Huilin, Tseng H W. Adaptive method to locate seed points based on information entropy and quadtree[J]. Sensors and Materials, 2021, 33(2): 789−804. doi: 10.18494/SAM.2021.3047 -
计量
- 文章访问数: 64
- HTML全文浏览量: 19
- PDF下载量: 11
- 被引次数: 0
