-
本文首次针对HL-3装置，采用机器学习方法预测偏滤器靶板等离子体参数，为未来快速预测大型聚变堆偏滤器热负荷奠定基础。将机器学习应用于边缘等离子体物理中，可以显著缩短大型边缘程序SOLPS-ITER模拟所需的时间，从几周、几个月甚至半年缩短至毫秒级。研究发现，通过增加内外偏滤器区的辐射损失作为模型的输入参数，能够明显提高预测精度（超过90%），同时增强训练模型的适用性，可以同时精确预测内外偏滤器靶板热流，并验证了该特征参数与偏滤器靶板物理量之间的依赖关系。该工作不仅为偏滤器物理研究提供了有效的方法，也为未来跨装置预测偏滤器靶板参数提供了坚实的基础。
-
关键词:
- HL-3 /
- 机器学习 /
- 神经网络 /
- 偏滤器靶板热负荷 /
- SOLPS-ITER
The SOLPS-ITER edge plasma simulation code has become a primary tool for divertor physics design and target plate heat load prediction in fusion research. However, SOLPS-ITER- based divertor design requires not only substantial computational time but also intensive hardware resources, which fundamentally limits its application in advancing divertor optimization, particularly in large-scale fusion reactor divertor design. In this paper, the machine learning method is used for the first time to predict the plasma parameters of the divertor target plate for HL-3, which provides a theoretical basis for predicting the heat load of divertor in large fusion reactor in the future. Based on the simulation of the edge plasma code SOLPS-ITER, we first build a database of HL-3 edge plasma parameters, including the upstream inner/outer midplane region and divertor target region. Then, we apply the machine learning method and combine with the database to develop an artificial neural network model. Finally, the artificial neural network is used to train a model using the boundary plasma parameters of the HL-3 device, and the heat load of the divertor target plate is predicted by the given upstream plasma parameters.
This work can effectively shorten the time for the edge plasma code SOLPS-ITER to simulate the edge plasma from weeks, months or even half a year to several ms. In this work, a multi-layer perceptron (MLP) model was established with different input parameters to predict the electron temperature, density and parallel heat flux of the inner and outer divertor target plates. It is found that reasonably increasing the upstream plasma parameters as the input to the model can not only enhance the generalization ability of the model and improve the accuracy of model prediction (both reaching more than 90%), but also verify the dependence between the upstream plasma parameters and the divertor target physical quantities. In addition, a more stable ResMLP model is established on the basis of MLP. This work proves the feasibility of using the neural network to predict the heat load of the divertor target plate.-
Keywords:
- HL-3 /
- machine learning /
- neural network /
- divertor target heat load /
- SOLPS-ITER
-
[1] 高翔, 万元熙, 丁宁, 彭先觉2018中国工程科学20 25
[2] Kuang A Q, Ballinger S, Brunner D, Canik J, Creely A J, Gray T, Greenwald M, Hughes J W, Irby J, LaBombard B, Lipschultz B, Lore J D, Reinke M L, Terry J L, Umansky M, Whyte D G, Wukitch S 2020 J. Plasma Phys. 86 865860505
[3] Du H L, Li J X, Xue L, Bonnin X, Zheng G Y, Xiao G L, Fan D M, Tong R H, Xue M, Song X, Wang S, Wu N, Ji X Q, Chen W, Zhong W L 2025 Nucl. Fusion 65 036023
[4] Shi B, Yang Z D, Zhang B, Yang C, Gan K F, Chen M W, Yang J H, Zhang H, Qi LJ L, Gong X Z, Zhang X D, Wang W H 2017Chin. Phys. Lett. 34 095201
[5] Pitts R A, Kukushkin A, Loarte A, Martin A, Merola M, Kessel C E, Komarov V, Shimada M 2009Phys. Scr. T138 014001
[6] Hayashi Y, Masuzaki S, Kobayashi M, Kawamura G, Mukai K, Tanaka H, Murase T 2025Plasma Phys. Control. Fusion 67 025010
[7] 杜海龙, 桑超峰, 王亮, 孙继忠, 刘少承, 汪惠乾, 张凌, 郭后扬, 王德真2013 62 245206
[8] Chen W T, Sun J Z, Gao F, Peng L, Wang D Z 2022 Chin. Phys. B 31 075204
[9] Wang M, Nie Q Y, Huang T, Wang X G, Zhang Y J 2024 Chin. Phys. B 33 035204
[10] Si H, Ding R, Senichenkov I, Rozhansky V, Molchanov P, Liu X J, Jia G Z, Sang C F, Mao S F, Chan V 2022Nucl. Fusion 62026031
[11] Peng L, Sun Z, Sun J Z, Maingi R, Gao F, Bonnin X, Chang H Y, Wang W K, Liu J Y 2024Chin. Phys. B 33 115201
[12] Moscheni M, Meineri C, Wigram M, Carati C, De Marchi E, Greenwald M, Innocente P, LaBombard B, Subba F, Wu H, Zanino R 2022Nucl. Fusion 62 056009
[13] Coster D P 2016Contrib. Plasma Phys 56 790
[14] 马锐垚, 王鑫, 李树, 勇珩, 上官丹骅2024 73 072802
[15] Li H, Fu Y L, Li J Q, Wang Z X 2023Chin. Phys. Lett 40125201
[16] 陈俊杰, 胡文慧, 肖建元, 郭笔豪, 肖炳甲2020计算机系统应用29 21
[17] Liu B, Wang Y X 2024Chin. Phys. B 33 084401
[18] Yan Z P, Wang Q Y, Yu X Q, Li J Z 2024 Chin. Phys. Lett 41 098901
[19] Yue J C, Chen R K, Ma D K, Hu S Q 2025Chin. Phys. Lett 42036301
[20] Bonnin X, Dekeyser W, Pitts R, Coster D, Voskoboynikov S, Wiesen S 2016Plasma Fusion Res. 11 1403102
[21] Schneider R, Bonnin X, Borrass K, Coster D P, Kastelewicz H, Reiter D, Rozhansky V A, Braams B J 2006Contrib. Plasma Phys. 46 3
[22] Sytova E, Senichenkov I, Kaveeva E, Rozhansky V, Veselova I, Voskoboynikov S, Coster D 201643rd EPS Conference on Plasma Physics Liege, Belgium, July 4–8, 2016 P1.054
[23] Stangeby P C, Sang C F 2017Nucl. Fusion 57 056007
[24] 马逐曦2023工业场景下铝板带表面缺陷的多尺度深度学习视觉检测方法研究博士学位论文(长沙:中南大学)
[25] Meha D, Manan S 2021Clin. eHealth 4 1
[26] Naskath J, Sivakamasundari G, Begum A A S 2022Wirel. Pers. Commun. 128 2913
[27] Glorot X, Bordes A, Bengio Y 2011The 14th International Conference on Artificial Intelligence and Statistics Fort Lauderdale, FL, USA, April 11–13, 2011 p315
[28] Nair V, Hinton G E 2010The 27th International Conference on Machine Learning (ICML) Haifa, Israel, June 21--24, 2010 P807
[29] Hochreiter S 1998Int. J. Uncertain. Fuzziness Knowl.-Based Syst. 6 107
[30] Krizhevsky A, Sutskever I, Hinton G E 2017Commun. ACM 6084
[31] Ren J J, Wang N 2018Gansu Gaoshi Xuebao 23 61
[32] Touvron H, Bojanowski P, Caron M, Cord M, El-Nouby A, Grave E, Izacard G, Joulin A, Synnaeve G, Verbeek J, Jégou H 2022IEEE Trans. Pattern Anal. Mach. Intell. 45 5314.
计量
- 文章访问数: 46
- PDF下载量: 0
- 被引次数: 0
下载:
