Study on multi-scale visual analysis method of activated corrosion products in fusion reactor

doi:10.11996/JG.j.2095-302X.2023030588

Abstract

Abstract:

Fusion energy is regarded as the ultimate clean energy for human use, and safety is paramount for the advancement of fusion energy. Experts in this field must ensure the safety of the entire process from production to usage. Currently, the problem of activated corrosion products in the cooling pipes of fusion reactors poses a significant challenge to safety. These products are the primary sources of radioactivity in fusion reactors and cause safety problems. They have a critical impact on the shielding design, personnel protection, and accident consequences of fusion reactors. Moreover, the activated corrosion products are distributed throughout the cooling pipe and accumulate over time. Therefore, studying their distribution characteristics in cooling pipes is essential for ensuring fusion safety. At this stage, experts in the field lack a method that can quickly and efficiently analyze the distribution characteristics of the activated corrosion products. During the research process, various parts of the cooling pipe of the fusion reactor were classified according to different characteristics, and a large number of components were obtained. Therefore, the activated corrosion products are widely distributed in the components and are composed of various radioactive substances such as Co57, Co60, and Mn54. The composition and distribution of the activated corrosion products evolve over time; therefore, the activated corrosion product data are multivariate time-series data. Considering the characteristics of the activated corrosion product data and the analytical needs of domain experts, a visual analysis system was designed and developed. Its main features include the following: ① To address the problem of too many components of the cooling pipe, it proposed a multi-level clustering analysis approach to divide the components and help domain experts quickly select representative components for in-depth analysis, thereby avoiding checking all components one by one. ② The system addressed the challenge of a large time span by automatically extracting potential critical time periods using the multi-granularity division method, thus enabling domain experts to focus on critical time periods. ③ A set of multi view linkage visual analysis systems was designed to assist domain experts in quickly obtaining the overall distribution characteristics of activated corrosion products through flexible interaction. Based on the above characteristics, a Web-based system that can conduct multi-scale and multi-granularity analyses of fusion reactor activated corrosion products was designed. The International Thermonuclear Experimental Reactor (ITER) is currently the largest international cooperation project in the world. The effectiveness of the method was tested using the activated corrosion product data of the ITER reactor from 2013 to 2020. Domain experts pointed out that the tool is useful in selecting representative components and key time periods and can greatly improve their work efficiency.

Key words: visibility analysis, multivariable time series data, multi-layer analysis, fusion nuclear safety, activated corrosion products

CLC Number:

TP391

LUO Yue-tong, YANG Meng-nan, PENG Jun, ZHOU Bo, ZHANG Yan-kong. Study on multi-scale visual analysis method of activated corrosion products in fusion reactor[J]. Journal of Graphics, 2023, 44(3): 588-598.

Figures/Tables 14

Fig. 1 Domain scientists analyze the process of activated corrosion products

Fig. 2 Data structure diagram

Fig. 3 System flow chart

Fig. 4 System interface ((a) Attribute configuration selection; (b) Cluster result analysis; (c) Parameter panel; (d) Component physical structure display; (e) Component association visualization; (f) Component specific time period data visualization)

Fig. 5 Chord diagram

Fig. 6 Node diagram

Fig. 7 Multilevel grouping operation

Fig. 8 Keyframe sequence

Fig. 9 Components search

Fig. 10 Multilevel grouping operation ((a) The first multilevel operation; (b) Last multilevel operation)

Fig. 11 Final clustering results

Fig. 12 Cluster expansion diagram

Fig. 13 Line chart of spatiotemporal data of representative components ((a) Comparison of different clusters; (b) Same cluster comparison)

Fig. 14 Abnormal component data characteristics

References 23

[1]	中国国际核聚变能源计划执行中心. “人造太阳”计划: 和平利用聚变能[J]. 国际人才交流, 2019(9): 8-10.
	China International Nuclear Fusion Energy Program Execution Center. “Artificial Sun” project: peaceful utilization of fusion energy[J]. International Talent, 2019(9): 8-10. (in Chinese)
[2]	张竞宇, 李璐, 宋文, 等. 水冷聚变堆活化腐蚀产物源项分析程序开发[J]. 原子能科学技术, 2015, 49(S1): 68-74.
	ZHANG J Y, LI L, SONG W, et al. Program development for source term analysis of activated corrosion product in water-cooled fusion reactor[J]. Atomic Energy Science and Technology, 2015, 49(S1): 68-74. (in Chinese)
[3]	DI PACE L, DACQUAIT F, SCHINDLER P, et al. Development of the PACTITER code and its application to safety analyses of ITER primary cooling water system[J]. Fusion Engineering and Design, 2007, 82(3): 237-247. DOI URL
[4]	ARDITSAS P J. Activation product transport using TRACT: ore estimation of an ITER cooling loop[J]. Fusion Engineering and Design, 1999, 45(2): 169-185. DOI URL
[5]	WU Y C, XIE X, WANG J C, et al. ForVizor: visualizing spatio-temporal team formations in soccer[J]. IEEE Transactions on Visualization and Computer Graphics, 2018, 25(1): 65-75 DOI URL
[6]	FUJIWARA T, SHILPIKA, SAKAMOTO N, et al. A visual analytics framework for reviewing multivariate time-series data with dimensionality reduction[J]. IEEE Transactions on Visualization and Computer Graphics, 2021, 27(2): 1601-1611. DOI URL
[7]	CHEN H C, YANG B, LIU J M, et al. Mining spatiotemporal diffusion network: a new framework of active surveillance planning[J]. IEEE Access, 2019, 7: 108458-108473. DOI URL
[8]	WANG X M, CHEN W, XIA J Z, et al. ConceptExplorer: visual analysis of concept drifts in multi-source time-series data[C]// 2020 IEEE Conference on Visual Analytics Science and Technology. New York: IEEE Press, 2021: 1-11.
[9]	AIGNER W, MIKSCH S, SCHUMANN H, et al. Time & time-oriented data[M]//Visualization of Time-Oriented Data. London: Springer London, 2011: 45-68.
[10]	HOCHHEISER H, SHNEIDERMAN B. Dynamic query tools for time series data sets: timebox widgets for interactive exploration[J]. Information Visualization, 2004, 3(1): 1-18. DOI URL
[11]	PENG R D. A method for visualizing multivariate time series data[J]. Journal of Statistical Software, 2008, 25(Code Snippet 1): 1-17.
[12]	PYLVÄNEN M, ÄYRÄMÖ S, KÄRKKÄINEN T. Visualizing time series state changes with prototype based clustering[M]// Adaptive and Natural Computing Algorithms. Heidelberg: Springer, 2009: 619-628.
[13]	WANG T D, PLAISANT C, SHNEIDERMAN B, et al. Temporal summaries: supporting temporal categorical searching, aggregation and comparison[J]. IEEE Transactions on Visualization and Computer Graphics, 2009, 15(6): 1049-1056. DOI PMID
[14]	BACH B, SHI C L, HEULOT N, et al. Time curves: folding time to visualize patterns of temporal evolution in data[J]. IEEE Transactions on Visualization and Computer Graphics, 2016, 22(1): 559-568. DOI URL
[15]	GU Y, WANG C L. TransGraph: hierarchical exploration of transition relationships in time-varying volumetric data[J]. IEEE Transactions on Visualization and Computer Graphics, 2011, 17(12): 2015-2024. DOI URL
[16]	LIU S X, YIN J L, WANG X T, et al. Online visual analytics of text streams[J]. IEEE Transactions on Visualization and Computer Graphics, 2016, 22(11): 2451-2466. DOI URL
[17]	LIU D Y, XU P P, REN L. TPFlow: progressive partition and multidimensional pattern extraction for large-scale spatio-temporal data analysis[J]. IEEE Transactions on Visualization and Computer Graphics, 2019, 25(1): 1-11.
[18]	ADHIKARI R, AGRAWAL R K. A combination of artificial neural network and random walk models for financial time series forecasting[J]. Neural Computing and Applications, 2014, 24(6): 1441-1449. DOI URL
[19]	BERTHOLD M R, HÖPPNER F. On clustering time series using euclidean distance and Pearson correlation[EB/OL]. [2022-02-13]. https://arxiv.org/abs/1601.02213.
[20]	BAI S H, QI H D, XIU N H. Constrained best euclidean distance embedding on a sphere: a matrix optimization approach[J]. SIAM Journal on Optimization, 2015, 25(1): 439-467. DOI URL
[21]	HSU C J, HUANG K S, YANG C B, et al. Flexible dynamic time warping for time series classification[J]. Procedia Computer Science, 2015, 51(1): 2838-2842. DOI URL
[22]	AßFALG J, KRIEGEL H P, KRÖGER P, et al. Similarity search on time series based on threshold queries[M]//Lecture Notes in Computer Science. Berlin, Heidelberg: Springer, 2006: 276-294.
[23]	DING H, TRAJCEVSKI G, SCHEUERMANN P, et al. Querying and mining of time series data[J]. Proceedings of the VLDB Endowment, 2008, 1(2): 1542-1552. DOI URL