Qingquan Yang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Qingquan Yang has a total of 609 co-authors. Here are the top co-authors by number of shared publications:
- Guo Sheng Xu
Chinese Academy of Sciences
34 shared publications - Xin Lin
Institute of Plasma Physics, Chinese Academy of Sciences, Hefei, Anhui 230031, China
27 shared publications - Liang Wang
Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031, People’s Republic of China
26 shared publications - Lingyi Meng
Chinese Academy of Sciences Institute of Plasma Physics
24 shared publications - Yi-Feng Wang
Chinese Academy of Sciences Institute of Plasma Physics
23 shared publications - Ning Yan
Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031, China
20 shared publications
Latest Publications
- 14 November 2026
- Published by Springer Nature
- p. 83-103
Temperature Field Reconstruction of Tungsten Monoblock Divertor on EAST Using Physics-Aware Neural Operator Transformer
- 14 November 2026
- Published by Springer Nature
- p. 83-103
Temperature Field Reconstruction of Tungsten Monoblock Divertor on EAST Using Physics-Aware Neural Operator Transformer
- 1 November 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (11) , 115016
Effects of strike point location on upstream electron density in HL-3 by SOLPS-ITER modeling
Experiments in L-mode on the HL-3 tokamak show that, under similar discharge conditions, the divertor strike point (SP) location significantly affects upstream edge and divertor plasma parameters. When the SP is located at the corner rather than the bottom, the upstream electron density (at the outer midplane n e,OMP ) is substantially higher. In this work, simulations were conducted using the edge plasma code SOLPS-ITER with the experimental parameters, which well reproduce the experimentally observed trends and provide insight into the mechanisms by which the SP location influences n e,OMP . The results show that when the SP is located at the corner, D atoms reflected from the right baffle accumulate in the downstream divertor region, enhancing the flux of D atoms entering the core region near the X -point. The subsequent ionization of these D atoms produces a stronger particle source in the core region and contributes to the higher core plasma density in the SPC case. Moreover, the simulations indicate that the difference in n e,sep between the SPC and SPB cases increases with input power. Besides stronger core ionization enhancing core plasma density in the SPC case, two additional effects contribute: (1) the peak D + flux incident on the corner divertor target ( ϕ D + ,OT peak ) sputters more C impurities, which are then transported upstream, resulting in a higher upstream carbon density n C,OMP and consequently increasing n e,OMP ; (2) in the SPC case, the D + poloidal flux through the outer divertor entrance to the downstream region is reduced, further increasing the difference in n e,OMP .
- 1 November 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (11) , 115016
Effects of strike point location on upstream electron density in HL-3 by SOLPS-ITER modeling
Experiments in L-mode on the HL-3 tokamak show that, under similar discharge conditions, the divertor strike point (SP) location significantly affects upstream edge and divertor plasma parameters. When the SP is located at the corner rather than the bottom, the upstream electron density (at the outer midplane n e,OMP ) is substantially higher. In this work, simulations were conducted using the edge plasma code SOLPS-ITER with the experimental parameters, which well reproduce the experimentally observed trends and provide insight into the mechanisms by which the SP location influences n e,OMP . The results show that when the SP is located at the corner, D atoms reflected from the right baffle accumulate in the downstream divertor region, enhancing the flux of D atoms entering the core region near the X -point. The subsequent ionization of these D atoms produces a stronger particle source in the core region and contributes to the higher core plasma density in the SPC case. Moreover, the simulations indicate that the difference in n e,sep between the SPC and SPB cases increases with input power. Besides stronger core ionization enhancing core plasma density in the SPC case, two additional effects contribute: (1) the peak D + flux incident on the corner divertor target ( ϕ D + ,OT peak ) sputters more C impurities, which are then transported upstream, resulting in a higher upstream carbon density n C,OMP and consequently increasing n e,OMP ; (2) in the SPC case, the D + poloidal flux through the outer divertor entrance to the downstream region is reduced, further increasing the difference in n e,OMP .
- Open Access
- 8 September 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (10) , 106041
Simulation study of tungsten erosion and impurity transport under the assumption of a full-tungsten wall in EAST plasmas
Future fusion reactors are expected to employ full-tungsten plasma-facing components (PFCs), and the generation and core accumulation of tungsten impurities under full-tungsten wall environment is one of the key issues that limit the stable operation of tokamak devices. In this work, the upgraded impurity transport code IMPEDGE is coupled with the extended-grid version of SOLPS-ITER to investigate the erosion and transport of tungsten impurity in full-tungsten wall environment. The background plasma in our simulation is based on experimental data from EAST discharge #140718, with all PFCs assumed to be tungsten. The simulations show that, under full-tungsten wall assumption, both the tungsten impurity erosion and its core concentration remain at relatively high levels even in the detached regime, with dominant contributions originating from the far scrape-off layer (far-SOL) of the lower divertor and the upper divertor region. The main reason is that, although the electron temperature at the near SOL (near-SOL) of the lower divertor decreases significantly during detachment, that in the far-SOL region remains difficult to further reduce (∼10 eV), resulting in substantial tungsten erosion. The total impurity penetration probability (defined as the probability for emitted particles to enter the plasma core) also remains at a relatively high level, mainly because the lower plasma density in the far-SOL weakens the impurity screening effect. In the downstream region, the reduced ion density decreases the friction force acting on impurities, allowing them to more easily escape from the region and migrate upstream. In the upstream region, the lower electron density increases the ionization mean free path, causing impurities to be ionized closer to the last closed flux surface and thereby enhancing impurity penetration. These results indicate that impurity screening in the near-SOL divertor region can be significantly enhanced by impurity-seeded detachment. However, erosion and penetration from the far-SOL region still dominate the core impurity accumulation, suggesting that more effective impurity control strategies for the far-SOL region are required in future studies.
- 8 September 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (10) , 106041
Open AccessSimulation study of tungsten erosion and impurity transport under the assumption of a full-tungsten wall in EAST plasmas
Future fusion reactors are expected to employ full-tungsten plasma-facing components (PFCs), and the generation and core accumulation of tungsten impurities under full-tungsten wall environment is one of the key issues that limit the stable operation of tokamak devices. In this work, the upgraded impurity transport code IMPEDGE is coupled with the extended-grid version of SOLPS-ITER to investigate the erosion and transport of tungsten impurity in full-tungsten wall environment. The background plasma in our simulation is based on experimental data from EAST discharge #140718, with all PFCs assumed to be tungsten. The simulations show that, under full-tungsten wall assumption, both the tungsten impurity erosion and its core concentration remain at relatively high levels even in the detached regime, with dominant contributions originating from the far scrape-off layer (far-SOL) of the lower divertor and the upper divertor region. The main reason is that, although the electron temperature at the near SOL (near-SOL) of the lower divertor decreases significantly during detachment, that in the far-SOL region remains difficult to further reduce (∼10 eV), resulting in substantial tungsten erosion. The total impurity penetration probability (defined as the probability for emitted particles to enter the plasma core) also remains at a relatively high level, mainly because the lower plasma density in the far-SOL weakens the impurity screening effect. In the downstream region, the reduced ion density decreases the friction force acting on impurities, allowing them to more easily escape from the region and migrate upstream. In the upstream region, the lower electron density increases the ionization mean free path, causing impurities to be ionized closer to the last closed flux surface and thereby enhancing impurity penetration. These results indicate that impurity screening in the near-SOL divertor region can be significantly enhanced by impurity-seeded detachment. However, erosion and penetration from the far-SOL region still dominate the core impurity accumulation, suggesting that more effective impurity control strategies for the far-SOL region are required in future studies.
- 26 June 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (26) , 265601
Direct classification and estimation of edge coherence mode parameters in EAST via attention-driven transformer networks
Edge coherent modes (ECMs) play a crucial role in regulating edge localized modes (ELMs) and facilitating the steady-state operation of H-mode plasmas. In this study, an end-to-end multi-task deep learning model based on the Transformer architecture, termed IFormer-ECM, has been developed to achieve direct identification and physical characterization of ECMs in EAST plasmas. By directly processing raw time-series signals from helium beam emission spectroscopy diagnostics, the model circumvents the need for tedious spectral calculations and manual parameter tuning. Experimental results demonstrate that the proposed model exhibits exceptional performance in the binary classification task of ECM detection, achieving an accuracy of 96 .45 % and a precision of 96 .00 % on the test dataset. Furthermore, the model is capable of simultaneously regressing key physical parameters of ECMs; the R 2 score for relative amplitude prediction reaches 0 .9832 , while the predictions for the frequency bounds (maximum and minimum frequencies) achieve R 2 scores of 0 .8855 and 0.9146, respectively. Additionally, an in-depth verification of the internal physical consistency among the predicted amplitude, f min and f max was rigorously performed utilizing Parseval’s theorem. This work demonstrates the significant potential of pure attention-based Transformer models in handling complex edge plasma instabilities, laying a solid technological foundation for real-time monitoring and active ELM control in future fusion devices.
- 26 June 2026
- Published by IOP Publishing in Physica Scripta
- Vol. 101 (26) , 265601
Direct classification and estimation of edge coherence mode parameters in EAST via attention-driven transformer networks
Edge coherent modes (ECMs) play a crucial role in regulating edge localized modes (ELMs) and facilitating the steady-state operation of H-mode plasmas. In this study, an end-to-end multi-task deep learning model based on the Transformer architecture, termed IFormer-ECM, has been developed to achieve direct identification and physical characterization of ECMs in EAST plasmas. By directly processing raw time-series signals from helium beam emission spectroscopy diagnostics, the model circumvents the need for tedious spectral calculations and manual parameter tuning. Experimental results demonstrate that the proposed model exhibits exceptional performance in the binary classification task of ECM detection, achieving an accuracy of 96 .45 % and a precision of 96 .00 % on the test dataset. Furthermore, the model is capable of simultaneously regressing key physical parameters of ECMs; the R 2 score for relative amplitude prediction reaches 0 .9832 , while the predictions for the frequency bounds (maximum and minimum frequencies) achieve R 2 scores of 0 .8855 and 0.9146, respectively. Additionally, an in-depth verification of the internal physical consistency among the predicted amplitude, f min and f max was rigorously performed utilizing Parseval’s theorem. This work demonstrates the significant potential of pure attention-based Transformer models in handling complex edge plasma instabilities, laying a solid technological foundation for real-time monitoring and active ELM control in future fusion devices.
- Open Access
- 24 June 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (8) , 086047
Impact of tungsten plasma facing components on H-mode operational space in EAST in support of ITER new baseline
Recent experiments in EAST have demonstrated the compatibility of tungsten (W) as main wall plasma facing material with high confinement H-mode with low or no boron coverage. The experiments were conducted in plasmas with q 95 ~ 6.0,which allowed access to both type-I and type-II ELMy H-modes in EAST with a W wall and low normalized input torque similar to ITER. Central Electron Cyclotron (EC), Neutral Beam Injection (NBI) and Lower Hybrid Waves (LHW) were applied as auxiliary heating in a range of total power injected into the plasma from 3 MW to 5 MW. Small amplitude high frequency type-II ELMs allow maintaining good H-mode energy confinement even when the distance between the separatrix and the main W limiter is as low as 4 cm. The normalized H-mode energy confinement can reach H 98 factors up to 1.1 for both EC+NBI and EC+LHW power combination in the type-II ELMy H-mode regime. In addition, for these type-II ELMy H-mode conditions, nitrogen puffing from divertor region has been used to achieve partial detachment without significant impact on H98 nor on the core W concentration.On the contrary, EAST operation in the type-I ELMy H-mode is strongly affected by the main wall W source. The use of n = 2 resonant magnetic perturbations (RMPs) to achieve type-I ELM suppression in EAST reduces the core W level but at the cost of about 10% reduction on energy confinement. These results complement existing W-wall findings, providing a preliminary foundation from EAST for evaluating the impact of the W first wall in ITER, as proposed in the new ITER baseline, and highlight that the achievement of sufficient ELM control levels with low deterioration of energy confinement is key to minimize this impact.
- 24 June 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (8) , 086047
Open AccessImpact of tungsten plasma facing components on H-mode operational space in EAST in support of ITER new baseline
Recent experiments in EAST have demonstrated the compatibility of tungsten (W) as main wall plasma facing material with high confinement H-mode with low or no boron coverage. The experiments were conducted in plasmas with q 95 ~ 6.0,which allowed access to both type-I and type-II ELMy H-modes in EAST with a W wall and low normalized input torque similar to ITER. Central Electron Cyclotron (EC), Neutral Beam Injection (NBI) and Lower Hybrid Waves (LHW) were applied as auxiliary heating in a range of total power injected into the plasma from 3 MW to 5 MW. Small amplitude high frequency type-II ELMs allow maintaining good H-mode energy confinement even when the distance between the separatrix and the main W limiter is as low as 4 cm. The normalized H-mode energy confinement can reach H 98 factors up to 1.1 for both EC+NBI and EC+LHW power combination in the type-II ELMy H-mode regime. In addition, for these type-II ELMy H-mode conditions, nitrogen puffing from divertor region has been used to achieve partial detachment without significant impact on H98 nor on the core W concentration.On the contrary, EAST operation in the type-I ELMy H-mode is strongly affected by the main wall W source. The use of n = 2 resonant magnetic perturbations (RMPs) to achieve type-I ELM suppression in EAST reduces the core W level but at the cost of about 10% reduction on energy confinement. These results complement existing W-wall findings, providing a preliminary foundation from EAST for evaluating the impact of the W first wall in ITER, as proposed in the new ITER baseline, and highlight that the achievement of sufficient ELM control levels with low deterioration of energy confinement is key to minimize this impact.
- 11 May 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (5)
Impact of electron inertia on edge localized mode simulation using BOUT++ six-field two-fluid model
The impact of electron inertia on the edge localized mode (ELM) simulations using the BOUT++ six-field two-fluid model is investigated in this work. The electron inertia is found to have negligible effect on the simulating linear growth rates under the typical tokamak edge plasma condition. Further analysis based on the ideal magnetohydrodynamics (MHD) model indicates that, the destabilizing effect of the electron inertia (as a high-order term) could be overestimated unless the related low-order terms, such as electron pressure gradient term, is considered in prior. For nonlinear simulations with the toroidally axisymmetric parallel current evolution, by including the electron inertia, the ELM burst can be successfully simulated without additional hyper-resistivity. The simulated ELM size lies between those simulated with constant hyper-resistivity of 10 -14 and 10 -13 . Radial spectrum analysis demonstrates that, during the nonlinear phase when the high-wavenumber components regarding the formation of narrow current sheets become considerable, the effect of electron inertia is non-negligible. It is found that the dominating dissipation mechanism is the electron momentum convection and the effect can be represented by the spatiotemporally varying effective hyper-resistivity with the highest value in the range between 10 -14 and 10 -13 . Furthermore, it is also found that the computational efficiency during the nonlinear phase can be improved when electron inertia is included.
- 11 May 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (5)
Impact of electron inertia on edge localized mode simulation using BOUT++ six-field two-fluid model
The impact of electron inertia on the edge localized mode (ELM) simulations using the BOUT++ six-field two-fluid model is investigated in this work. The electron inertia is found to have negligible effect on the simulating linear growth rates under the typical tokamak edge plasma condition. Further analysis based on the ideal magnetohydrodynamics (MHD) model indicates that, the destabilizing effect of the electron inertia (as a high-order term) could be overestimated unless the related low-order terms, such as electron pressure gradient term, is considered in prior. For nonlinear simulations with the toroidally axisymmetric parallel current evolution, by including the electron inertia, the ELM burst can be successfully simulated without additional hyper-resistivity. The simulated ELM size lies between those simulated with constant hyper-resistivity of 10 -14 and 10 -13 . Radial spectrum analysis demonstrates that, during the nonlinear phase when the high-wavenumber components regarding the formation of narrow current sheets become considerable, the effect of electron inertia is non-negligible. It is found that the dominating dissipation mechanism is the electron momentum convection and the effect can be represented by the spatiotemporally varying effective hyper-resistivity with the highest value in the range between 10 -14 and 10 -13 . Furthermore, it is also found that the computational efficiency during the nonlinear phase can be improved when electron inertia is included.
- Open Access
- 16 March 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (4) , 044002
Simultaneous ELM suppression and divertor detachment via synergistic boron powder and neon injection in EAST
A novel approach for simultaneous power exhaust and edge-localized mode (ELM) control is presented in the Experimental Advanced Superconducting Tokamak discharges, which utilize an ITER-like tungsten divertor. Real-time injection of boron (B) powder and neon (Ne) gas overcomes their limitations encountered when used separately. Pure Ne seeding leads to a narrow operational window constrained by core impurity accumulation and H-mode to L-mode back transitions, while pure solid B injection (SBI) is insufficient for effective divertor cooling. In comparison, their combined use achieves a stable, stationary, ELM-suppressed H-mode with adequate power exhaust. This synergistic scenario features partial energy detachment at the outer divertor while maintaining good plasma confinement ( H 98 ∼ 1) with minimal degradation. Two key features of this scenario are: (1) the SBI triggers a persistent Edge Harmonic Mode (EHM), which provides a crucial continuous particle transport channel, preventing Ne and tungsten/molybdenum accumulation without flushing out by ELM, and (2) the B + Ne mixture allows for active optimization of the radiated power profile. Core radiation can be reduced by substituting a portion of the Ne with B, leveraging their complementary non-coronal equilibrium radiation efficiencies. This combined B + Ne injection scheme presents a promising pathway toward integrated core-edge scenarios, offering the potential to minimize total impurity throughput while leveraging an actuator (powder injection) already being considered for ITER.
- 16 March 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (4) , 044002
Open AccessSimultaneous ELM suppression and divertor detachment via synergistic boron powder and neon injection in EAST
A novel approach for simultaneous power exhaust and edge-localized mode (ELM) control is presented in the Experimental Advanced Superconducting Tokamak discharges, which utilize an ITER-like tungsten divertor. Real-time injection of boron (B) powder and neon (Ne) gas overcomes their limitations encountered when used separately. Pure Ne seeding leads to a narrow operational window constrained by core impurity accumulation and H-mode to L-mode back transitions, while pure solid B injection (SBI) is insufficient for effective divertor cooling. In comparison, their combined use achieves a stable, stationary, ELM-suppressed H-mode with adequate power exhaust. This synergistic scenario features partial energy detachment at the outer divertor while maintaining good plasma confinement ( H 98 ∼ 1) with minimal degradation. Two key features of this scenario are: (1) the SBI triggers a persistent Edge Harmonic Mode (EHM), which provides a crucial continuous particle transport channel, preventing Ne and tungsten/molybdenum accumulation without flushing out by ELM, and (2) the B + Ne mixture allows for active optimization of the radiated power profile. Core radiation can be reduced by substituting a portion of the Ne with B, leveraging their complementary non-coronal equilibrium radiation efficiencies. This combined B + Ne injection scheme presents a promising pathway toward integrated core-edge scenarios, offering the potential to minimize total impurity throughput while leveraging an actuator (powder injection) already being considered for ITER.
- 9 March 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (3)
Spatiotemporal Characteristics of Type-I ELM Power Deposition Measured by Infrared Thermography on the EAST Tokamak
The spatiotemporal characteristics of natural and resonant magnetic perturbation (RMP)-mitigated type-I edge localized modes (ELMs) on EAST are investigated using a high-resolution infrared (IR) camera. Natural ELMs exhibit a rise time (τ_"rise" ) of 350–750 μs and a decay time (τ_decay) of 600–2300 μs, corresponding to τ_decay⁄τ_rise = 1.5–4. For RMP-mitigated ELMs, τ_"rise" and τ_decay are shorter than those of natural ELMs, which is correlated with a reduction in ELM size ((ΔW_ELM)/W_MHD ). Further analysis shows that, for both natural and RMP-mitigated ELMs, peak heat flux (q_peak) is largely independent of ELM size, whereas wetted area (A_wet) increases with ELM size. Moreover, RMP-mitigated ELMs exhibit slightly lower q_peak than natural ELMs, indicating a modest reduction in transient heat load even under weak mitigation. Comparison between measured peak energy fluence and Eich model predictions shows good agreement for both natural and RMP-mitigated ELMs, with a slightly lower peak energy fluence under RMP, likely due to reduced ELM size. These results suggest that RMP modulates both the temporal and spatial characteristics of ELMs on EAST, which could have implications for ELM control strategies.
- 9 March 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (3)
Spatiotemporal Characteristics of Type-I ELM Power Deposition Measured by Infrared Thermography on the EAST Tokamak
The spatiotemporal characteristics of natural and resonant magnetic perturbation (RMP)-mitigated type-I edge localized modes (ELMs) on EAST are investigated using a high-resolution infrared (IR) camera. Natural ELMs exhibit a rise time (τ_"rise" ) of 350–750 μs and a decay time (τ_decay) of 600–2300 μs, corresponding to τ_decay⁄τ_rise = 1.5–4. For RMP-mitigated ELMs, τ_"rise" and τ_decay are shorter than those of natural ELMs, which is correlated with a reduction in ELM size ((ΔW_ELM)/W_MHD ). Further analysis shows that, for both natural and RMP-mitigated ELMs, peak heat flux (q_peak) is largely independent of ELM size, whereas wetted area (A_wet) increases with ELM size. Moreover, RMP-mitigated ELMs exhibit slightly lower q_peak than natural ELMs, indicating a modest reduction in transient heat load even under weak mitigation. Comparison between measured peak energy fluence and Eich model predictions shows good agreement for both natural and RMP-mitigated ELMs, with a slightly lower peak energy fluence under RMP, likely due to reduced ELM size. These results suggest that RMP modulates both the temporal and spatial characteristics of ELMs on EAST, which could have implications for ELM control strategies.
- 1 March 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (3) , 035024
Data-driven prediction of the: L–H transition power threshold in the EAST tokamak using ensemble learning regression
Accurate prediction of the L–H transition power threshold is essential for achieving steady-state high-performance burning plasmas in future fusion reactors. In this study, we develop a machine learning–based regression framework to predict the plasma power loss across the separatrix, P loss , at the onset of the L–H transition. The model is trained on 3308 EAST discharges; from each discharge, we extract exactly one sample at the onset of its first L–H transition. Three ensemble learning models—random forest, XGBoost, and CatBoost—are systematically compared. As a benchmark, we compare our results with the empirical ITPA scaling law, which was derived from the high-density branch of the L–H power threshold in favorable ion ∇ B drift configurations of ITER-like plasmas. On our present EAST database, which spans both high-density and low-density regimes and includes discharges with both favorable and unfavorable ∇ B drift directions, the CatBoost model attains the highest predictive accuracy and yields significantly lower errors than the multivariate linear regression baseline. Model interpretation is performed using SHAP (SHapley Additive exPlanations). The SHAP rankings indicate that the key variables appearing in empirical L–H scaling relations ( S , n ¯ e , and B T ) remain the dominant predictors, while the plasma current I p is consistently identified as an additional leading contributor. Equilibrium-related and geometry-related parameters ( δ top , δ bot , q 95 , κ ) also exhibit non-negligible contributions. Underexplored spatial variables such as Z LX show a notable impact on the power threshold within the present EAST database. These results suggest that the machine-learning analysis can reproduce the main dependencies embodied in conventional scaling laws, while also indicating additional effects that are not explicitly encoded in the Martin scaling. The present framework provides an alternative data-driven approach for predicting the L–H transition, which may be particularly useful for exploring how multiple coupled control parameters jointly influence the L–H power threshold in complex, reactor-relevant operating scenarios.
- 1 March 2026
- Published by IOP Publishing in Plasma Physics and Controlled Fusion
- Vol. 68 (3) , 035024
Data-driven prediction of the: L–H transition power threshold in the EAST tokamak using ensemble learning regression
Accurate prediction of the L–H transition power threshold is essential for achieving steady-state high-performance burning plasmas in future fusion reactors. In this study, we develop a machine learning–based regression framework to predict the plasma power loss across the separatrix, P loss , at the onset of the L–H transition. The model is trained on 3308 EAST discharges; from each discharge, we extract exactly one sample at the onset of its first L–H transition. Three ensemble learning models—random forest, XGBoost, and CatBoost—are systematically compared. As a benchmark, we compare our results with the empirical ITPA scaling law, which was derived from the high-density branch of the L–H power threshold in favorable ion ∇ B drift configurations of ITER-like plasmas. On our present EAST database, which spans both high-density and low-density regimes and includes discharges with both favorable and unfavorable ∇ B drift directions, the CatBoost model attains the highest predictive accuracy and yields significantly lower errors than the multivariate linear regression baseline. Model interpretation is performed using SHAP (SHapley Additive exPlanations). The SHAP rankings indicate that the key variables appearing in empirical L–H scaling relations ( S , n ¯ e , and B T ) remain the dominant predictors, while the plasma current I p is consistently identified as an additional leading contributor. Equilibrium-related and geometry-related parameters ( δ top , δ bot , q 95 , κ ) also exhibit non-negligible contributions. Underexplored spatial variables such as Z LX show a notable impact on the power threshold within the present EAST database. These results suggest that the machine-learning analysis can reproduce the main dependencies embodied in conventional scaling laws, while also indicating additional effects that are not explicitly encoded in the Martin scaling. The present framework provides an alternative data-driven approach for predicting the L–H transition, which may be particularly useful for exploring how multiple coupled control parameters jointly influence the L–H power threshold in complex, reactor-relevant operating scenarios.
- Open Access
- 13 February 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (3) , 036028
Impact of toroidal magnetic field direction on integrated ELM-stable operation and divertor power exhaust via boron powder injection in EAST
We report the first in-depth comparison of the impact of toroidal magnetic field direction on solid boron injection used for Edge-Localized Mode (ELM) control, power exhaust, and core high- Z impurity control in the Experimental Advanced Superconducting Tokamak. With favorable ion ∇ B drift towards the upper X-point in an upper-single-null configuration, boron injection effectively suppresses ELMs, produces a detachment of the inner divertor target, and leads to improved energy confinement. ELM suppression in this configuration is accompanied by the excitation of an Edge Harmonic Mode. In contrast, with unfavorable ion ∇ B drift away from the upper X-point, boron injection also suppresses ELMs but leads to a more symmetric detachment state of both the inner and outer divertor targets, while plasma energy confinement is slightly degraded despite similar boron injection levels; a different low-frequency coherent mode without multiple harmonics is observed. Measurements from toroidally separated views show that the divertor response to boron injection is essentially toroidally symmetric, supporting the use of two-dimensional SOLPS-ITER modeling with a toroidally uniform impurity source. These experimental observations are qualitatively consistent with SOLPS-ITER simulations, which highlight the critical role of E × B drift effects in setting the Bt -dependent in–out asymmetry of detachment and in asymmetrically transporting particles and injected impurities within the scrape-off layer and private-flux region. These findings underscore the importance of drift physics and real-time wall conditioning in controlling low- Z impurity transport and optimizing edge solutions for integrated, ELM-stable, high-performance tokamak operation.
- 13 February 2026
- Published by IOP Publishing in Nuclear Fusion
- Vol. 66 (3) , 036028
Open AccessImpact of toroidal magnetic field direction on integrated ELM-stable operation and divertor power exhaust via boron powder injection in EAST
We report the first in-depth comparison of the impact of toroidal magnetic field direction on solid boron injection used for Edge-Localized Mode (ELM) control, power exhaust, and core high- Z impurity control in the Experimental Advanced Superconducting Tokamak. With favorable ion ∇ B drift towards the upper X-point in an upper-single-null configuration, boron injection effectively suppresses ELMs, produces a detachment of the inner divertor target, and leads to improved energy confinement. ELM suppression in this configuration is accompanied by the excitation of an Edge Harmonic Mode. In contrast, with unfavorable ion ∇ B drift away from the upper X-point, boron injection also suppresses ELMs but leads to a more symmetric detachment state of both the inner and outer divertor targets, while plasma energy confinement is slightly degraded despite similar boron injection levels; a different low-frequency coherent mode without multiple harmonics is observed. Measurements from toroidally separated views show that the divertor response to boron injection is essentially toroidally symmetric, supporting the use of two-dimensional SOLPS-ITER modeling with a toroidally uniform impurity source. These experimental observations are qualitatively consistent with SOLPS-ITER simulations, which highlight the critical role of E × B drift effects in setting the Bt -dependent in–out asymmetry of detachment and in asymmetrically transporting particles and injected impurities within the scrape-off layer and private-flux region. These findings underscore the importance of drift physics and real-time wall conditioning in controlling low- Z impurity transport and optimizing edge solutions for integrated, ELM-stable, high-performance tokamak operation.