XYXingting Yan

Xingting Yan

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xingting Yan has a total of 86 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • DKEKAN: a single-parameterized KAN surrogate for drift kinetic equation toward fast neoclassical toroidal viscosity torque modeling in tokamaks

    by Jinpeng Huang, Xingting Yan, Mingyu Zhang, Nana Bao, Zixuan Song, Yuetao Meng, Weiyong Zhou, Youwen Sun

    The neoclassical toroidal viscosity (NTV) torque is a critical driver of toroidal rotation in tokamaks, profoundly influencing plasma stability and performance. Although incorporating NTV effects is essential for integrated modeling, the high computational cost of NTV modeling has limited such self-consistent integration. This bottleneck arises because NTV calculation requires solving its governing equation—the drift kinetic equation (DKE)—in high-dimensional phase space. To address this issue, this study develops DKEKAN, a single-parameterized Kolmogorov–Arnold Network (SKAN) surrogate for solving DKE, to realize fast NTV modeling in tokamaks. With its physics-grouped expert layer and SKAN backbone, DKEKAN outperforms the tested multilayer perceptron, KAN, and neural-operator baselines in overall prediction accuracy, while reducing the standalone DKE-solving time from 35.85 s to 3.72 s, corresponding to a speedup of approximately 9.6 × , and reducing the total coupled NTVTOK runtime from 38.24 s to 5.58 s, corresponding to an overall speedup of approximately 6.9 × . This work effectively overcomes the computational bottleneck in NTV simulations, thus supporting further integrated modeling that incorporates NTV effects.

  • Open Access

    High-fidelity and explainable machine learning method for neoclassical toroidal viscosity torque prediction

    by Nana Bao, Weiyong Zhou, Xingting Yan, Youwen Sun, Yuetao Meng, Ziqi Lin, Jinpeng Huang, Yanxi Lu

    A high-fidelity and explainable machine learning (ML) method for neoclassical toroidal viscosity (NTV) torque prediction is investigated in this work. The ML surrogate model is trained based on NTV torque datasets obtained from first-principle numerical modelings, utilizing the extreme gradient boosting (XGBoost) algorithm and the Bayesian hyper-parameter optimization strategy. The trained model exhibits high statistical accuracy regarding the coefficient of determination R2, the mean squared error M SE and the relative error εs for NTV torque prediction, and about 2 orders of magnitude computational speed-up compared to physics-driven simulations. The prediction accuracy is further improved by applying K-fold cross-validation averaging method. In particular, the probability of large local error occurrence is reduced. Due to the radial locality of data samples, the trained model is able to make predictions for arbitrary numerical resolutions which can be different from the training data. Explainable artificial intelligence (XAI) technique, i.e. the Shapley additive explanations (SHAP) method is used to quantify the global and local contributions of input features to model output, which shows physically meaningful feature relationships, indicating alignments with the domain knowledge thus improving the model reliability. The fast, high-fidelity and explainable features of ML surrogate model makes self-consistent nonlinear simulations and other real-time applications possible for future tokamak physics studies.


  • Open Access

    Effects of neoclassical toroidal viscosity on toroidal rotation under electron cyclotron resonance heating on J-TEXT tokamak

    by Hanhui Li, Youwen Sun, Lu Wang, Xingting Yan, Hui Sheng, Zhoujun Yang, Li Gao, Yonghua Ding, Zhongyong Chen, Zhipeng Chen, et al.

    Influence of neoclassical toroidal viscosity (NTV) torque on intrinsic toroidal rotation caused by internal kink mode (IKM) in the J-TEXT tokamak is studied. It is observed that the toroidal rotation in the counter-current direction decreases after the on-axis electron cyclotron resonance heating (ECRH) is turned on, during which the IKM is strongly enhanced. The change of plasma rotation due to ECRH decreases with increasing plasma density, which is due to a decreasing of the IKM’s amplitude. This gives direct evidence that the change in intrinsic rotation strongly depends on the amplitude of the IKM. It is shown in the modeling that the decrease in the amplitude of the IKM reduces the NTV torque, and hence the influence on the toroidal rotation becomes weaker. The modeled dependence of rotation variation on plasma density agrees well with the observations. These results indicate that the NTV torque caused by the IKM plays a key role in changing the intrinsic core toroidal rotation in ECRH plasmas in the J-TEXT tokamak.


  • Open Access

    Neoclassical impurities transport induced by 3D fields

    by H. Sheng, Y. W. Sun, W. M. Zhang, C. Ye, Y. Q. Liu, X. T. Yan, S. F. Mao, H. H. Wang, L. Zhang, S. Gu, et al.

    The neoclassical impurities transport induced by 3D fields, known as the neoclassical toroidal viscosity (NTV) flux, is discussed in detail using NTVTOK. A substantial inward NTV flux of impurities is observed, which plays a key role in impurity transport and provides a mechanism for the enhancement of impurities intensity induced by resonant magnetic perturbations (RMP). The inward NTV impurities flux results from the negative gap between the E→×B→ drift frequency and the diamagnetic drift frequency of the impurities, a condition that can occur when the impurities gradient is positive. Consequently, this effect is strongly dependent on the spatial distribution of the impurities, in particular the direction and magnitude of the impurities gradient. It is also influenced by the strength of the 3D fields and can be controlled by changing the RMP configuration. The complex dependence of the NTV impurity flux on the toroidal rotation, mass, charge and ion temperature of impurities is also presented and discussed in detail. These factors will alter the collisionality of the impurities, the gap between the E→×B→ drift frequency and the diamagnetic drift frequency, the precession and the bounce-drift resonance, and consequently affect the NTV flux of the impurities. These results improve the understanding of RMP on impurities transport and provide a possible physical basis for the impurities control using RMP.


  • NTVTOK-ML: Fast surrogate model for neoclassical toroidal viscosity torque calculation in tokamaks based on machine learning methods

    by X.-T. Yan, N.-N. Bao, C.-Y. Zhao, Y.-W. Sun, Y.-T. Meng, W.-Y. Zhou, N.-Y. Liang, Y.-X. Lu, Y.-F. Liang, B.-N. Wan


  • Open Access

    Acceleration of plasma toroidal rotation driven by non-axisymmetric magnetic perturbation fields in the EAST tokamak

    by H. Sheng, B. Lyu, Y. W. Sun, H. H. Li, C. Bae, Y. Q. Liu, Y. F. Jin, S. F Mao, X. T. Yan, P. C. Xie, et al.

    Plasma toroidal rotation acceleration in the co-current direction introduced by the n = 1 (toroidal mode number) static resonant magnetic perturbation (RMP) has been observed in the EAST tokamak. It strongly depends on the RMP coil configuration, which is manifested by its dependence on δϕUL (phase difference between upper and lower coils) and RMP current. Modeling results from NTVTOK based on the linear plasma response modeled by MARS-F shows that the Neoclassical Toroidal Viscosity (NTV) torque is in the co-current direction because of the dominant contribution from electrons with the condition that the electron normalized collisionality is much lower than that of ions in this experiment. The modeled dependence of core integrated NTV torque modulated by the magnitude of core magnetic perturbation on δϕUL is consistent with the experimental observations. Threshold condition related to normalized collisionality to achieve the transition from rotation braking to acceleration is obtained in the NTV modeling and agrees well with experimental observations. It is shown in the modeling that the discharges with rotation acceleration are located at the regime that electron contribution to NTV is dominant and the torque is in co-current direction, while others with rotation braking are located at the regime that ion contribution to NTV torque is dominant and the torque is in countercurrent direction. Though the modeling results are in qualitative agreement with the experimental results, there is quantity difference between the modeled NTV torque based on linear plasma response and the experimental values. Possible reason is that the 3D fields are underestimated by linear modeling, particularly in the case of RMP field penetration, as demonstrated by the RMP current threshold for the rotation acceleration observed in the experiments.


  • Open Access

    Tungsten transport due to the neoclassical toroidal viscosity induced by resonant magnetic perturbation in the EAST tokamak

    by Youyou Chang, Youwen Sun, Hui Sheng, Xingting Yan, Hanhui Li, Pengcheng Xie, Wenmin Zhang, Jin Guo, Shifeng Mao, Yueqiang Liu, et al.

    The impact of neoclassical toroidal viscosity (NTV) induced by resonant magnetic perturbations (RMPs) on tungsten has been studied in EAST. Modeling results from NTVTOK show that the NTV effect on low charge state tungsten ions is greater than that on high charge state tungsten ions, and the effective transport coefficient of W20+ is only on the order of 10−4–10−3 m2/s, while the effective transport coefficient of W1+ can reach 0.1–1 m2/s, which is close to the 2D neoclassical diffusion coefficient of W1+. The flux of W1+ exhibits a clear phase dependence on the RMP spectrum, and the W1+ flux obtained from NTVTOK is higher at low toroidal mode numbers compared to high toroidal mode numbers of RMPs, which is qualitatively consistent with EAST experiments.


  • Open Access

    Implementing a new fully stepwise decomposition-based sampling technique for the hybrid water level forecasting model in real-world application

    by Ziqian Zhang, Nana Bao, Xingting Yan, Aokai Zhu, Chenyang Li, Mingyu Liu

    Various time variant non-stationary signals need to be pre-processed properly in hydrological time series forecasting in real world, for example, predictions of water level. Decomposition method is a good candidate and widely used in such a pre-processing problem. However, decomposition methods with an inappropriate sampling technique may introduce future data which is not available in practical applications, and result in incorrect decomposition-based forecasting models. In this work, a novel Fully Stepwise Decomposition-Based (FSDB) sampling technique is well designed for the decomposition-based forecasting model, strictly avoiding introducing future information. This sampling technique with decomposition methods, such as Variational Mode Decomposition (VMD) and Singular spectrum analysis (SSA), is applied to predict water level time series in three different stations of Guoyang and Chaohu basins in China. Results of VMD-based hybrid model using FSDB sampling technique show that Nash-Sutcliffe Efficiency (NSE) coefficient is increased by 6.4%, 28.8% and 7.0% in three stations respectively, compared with those obtained from the currently most advanced sampling technique. In the meantime, for series of SSA-based experiments, NSE is increased by 3.2%, 3.1% and 1.1% respectively. We conclude that the newly developed FSDB sampling technique can be used to enhance the performance of decomposition-based hybrid model in water level time series forecasting in real world.

  • Open Access

    Modeling of toroidal momentum transport induced by neoclassical toroidal viscosity torque for ITER scenarios

    by Xingting Yan, Youwen Sun, L. Li, Yueqiang Liu, N.-N. Bao, Alberto Loarte, Simon D Pinches, Baonian Wan

    Neoclassical toroidal viscosity (NTV) torque caused by resonant magnetic perturbation (RMP) and the induced toroidal momentum transport are investigated for International Thermonuclear Experimental Reactor (ITER) scenarios through numerical modeling. The NTV torque is calculated using the NTVTOK code including the bounce-drift resonant effect, and the toroidal rotation evolution is modeled by solving a toroidal momentum transport equation that couples momentum source (NTV torque) and the momentum diffusion effect. The variation of RMP coil phasing (defined as toroidal phase difference between different rows of RMP coils) results in different types of plasma response and hence different features of NTV torque and toroidal momentum transport. The bounce-drift resonant effect enhances NTV torque and induces more significant toroidal rotation variation than simulations that adopt the bounce-averaged NTV model. With the initial rotation of the ITER design, plasma rotation is braked by NTV torque, but it may be sustained at moderate amplitude due to electron contributions to NTV torque. It is also found that initially static or slowly rotating plasma can be accelerated by NTV torque either toward co- I p or counter- I p ( I p indicates plasma current) direction, indicating that NTV torque can be regarded as a momentum source for plasma with low torque injection; for instance, radio-frequency heated plasma.