AGAndreas Gillgren

Andreas Gillgren

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Andreas Gillgren has a total of 33 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Interpretability guided transfer learning approaches for tritium pedestal predictions

    by Andreas Gillgren, Dimitriy Yadykin, Pär Strand

    We explore transfer learning approaches to extend data-driven pedestal models trained on deuterium (D) plasmas to tritium (T) and DT mixtures. Specifically, we use models pre-trained on JET D pulses, and JET T/DT data for the transfer learning. We use model interpretability to guide our choice of transfer learning strategy. Analysis of model behavior post transfer learning reveals that sparsity and multicollinearity in the T/DT data lead to severe overfitting when fine-tuning the weights of the pre-trained neural network-based models. Therefore, we instead use a more robust and simple output calibration approach to facilitate the impact of isotope composition. This yields models with R2 between 0.66-0.87, performing significantly better than uncalibrated models though not matching the original D-only performance. The scaling coefficients obtained qualitatively agree with previous research, namely that the pedestal density scales positively with increased isotope mass, while pedestal temperature exhibits a weak negative scaling with isotope mass. This work highlights the importance of understanding model behavior in transfer learning to ensure reasonable functional mappings, which is particularly relevant for fusion research where sparse, multicollinear data are encountered.

  • Open Access

    Investigating characteristics of the growth rates from QuaLiKiz using an interpretable surrogate model

    by A. Gillgren, E. Fransson, A. Ludvig-Osipov, W. Enström, L. Flyckt, M. Green, M. Kvartsén, Y. Liljegren, E. Olsson, A. Orthag, et al.

    We present an interpretable, machine learning-based surrogate model for the eigenvalue solver in QuaLiKiz, a model that simulates turbulent transport in fusion plasmas. The aim is to exploit prediction transparency to gain insight into the anticipated behavior of QuaLiKiz-based surrogates and the underlying eigenvalue solver, a task that is more challenging when using black-box surrogate models. Specifically, we focus on predicting the growth rate of turbulence driving ion temperature gradient instabilities computed by QuaLiKiz for the normalized poloidal wavenumber kθρs=0.325. We split the task into a classification task, to determine whether the growth rate is positive (unstable mode) or not, and a growth rate prediction task, knowing the mode is unstable. The dataset used is a QuaLiKiz dataset based on JET pulses. The method used is the NeuralBranch method, a neural network-based method that reveals how the inputs of the models, in this case plasma parameters, impact the output. Results show that NeuralBranch models outperform linear models and match dense neural networks (traditional black-box models) in accuracy while being interpretable. By analyzing the NeuralBranch models, we identify parameter dependencies that cannot be captured by linear models. For instance, the models indicate that the stabilizing effect of ExB shear on the growth rate is suppressed at low magnetic shear, which can be attributed to how ExB shear influences the eigenfunction width in QuaLiKiz. In summary, this work demonstrates how interpretable methods can shed light on the behavior of surrogates and their underlying counterpart, thus enhancing both model credibility and understanding.


  • Open Access

    Investigating pedestal dependencies at JET using an interpretable neural network architecture

    by A. Gillgren, A. Ludvig-Osipov, D. Yadykin, P. Strand, Jet Contributors

    We present NeuralBranch, an interpretable neural network framework. In this work, we use it specifically to predict the pedestal from key engineering parameters in tokamak fusion experiments. The main goal is to uncover intricate relationships that traditional power scalings, with their limited expressive capacity, fail to capture. A secondary objective is to provide a transparent alternative to current opaque, black-box machine learning models used to predict the pedestal in integrated modeling frameworks. By using the proposed method, we obtain a novel global overview of several intricate dependencies in the JET pedestal database. For instance, while both input power and plasma current are positively correlated with pedestal top pressure and temperature, NeuralBranch reveals an attenuating interaction. This means that increasing power weakens the impact that current has on pedestal pressure and temperature, and vice versa. Further investigation of this interaction may be important to avoid overestimating pedestal stored energy at future machines like ITER when using established power scalings. We also identify an amplifying interaction between plasma current and triangularity, where higher triangularity amplifies the effect of plasma current on pedestal density, and vice versa. In addition to these findings, NeuralBranch matches the accuracy of black-box neural networks, with R 2 values as high as 0.88. This demonstrates that interpretability, with its associated benefits, can be achieved without sacrificing accuracy, making NeuralBranch a promising alternative for pedestal predictions.


  • Open Access

    Data-driven models in fusion exhaust: AI methods and perspectives

    by S. Wiesen, S. Dasbach, A. Kit, A.E. Jaervinen, A. Gillgren, A. Ho, A. Panera, D. Reiser, M. Brenzke, Y. Poels, et al.

    A review is given on the highlights of a scatter-shot approach of developing machine-learning methods and artificial neural networks based fast predictors for the application to fusion exhaust. The aim is to enable and facilitate optimized and improved modeling allowing more flexible integration of physics models in the light of extrapolations towards future fusion devices. The project encompasses various research objectives: (a) developments of surrogate model predictors for power & particle exhaust in fusion power plants; (b) assessments of surrogate models for time-dependent phenomena in the plasma-edge; (c) feasibility studies of micro–macro model discovery for plasma-facing components surface morphology & durability; and (d) enhancements of pedestal models & databases through interpolators and generators exploiting uncertainty quantification. Presented results demonstrate useful applications for machine-learning and artificial intelligence in fusion exhaust modeling schemes, enabling an unprecedented combination of both fast and accurate simulation.


  • High temporal resolution of pedestal dynamics via machine learning on density diagnostics

    by Diogo R Ferreira, Andreas Gillgren, Andrei Ludvig-Osipov, Pär Strand, Jet Contributors

    At the Joint European Torus, the reference diagnostic to measure electron density is Thomson scattering. However, this diagnostic has a low sampling rate, which makes it impractical to study the temporal dynamics of fast processes, such as edge localized modes. In this work, we use machine learning to predict the density profile based on data from another diagnostic, namely reflectometry. By learning to transform reflectometry data into Thomson scattering profiles, the model is able to generate the density profile at a much higher sampling rate than Thomson scattering, and more accurately than reflectometry alone. This enables the study of pedestal dynamics, by analyzing the time evolution of the pedestal height, width, position and gradient. We also discuss the accuracy of the model when applied on experimental campaigns that are different from the one it was trained on.


  • Open Access

    A fast neural network surrogate model for the eigenvalues of QuaLiKiz

    by E. Fransson, A. Gillgren, A. Ho, J. Borsander, O. Lindberg, W. Rieck, M. Åqvist, P. Strand

    We introduce a neural network surrogate model that predicts the eigenvalues for the turbulent microinstabilities, based on the gyrokinetic eigenvalue solver in QuaLiKiz. The model quickly provides information about the dominant instability for specific plasma conditions, and in addition, the eigenvalues offer a pathway for extrapolating transport fluxes. The model is trained on a 5 × 106 data points large dataset based on experimental data from discharges at the joint European torus, where each data point represents a QuaLiKiz simulation. The most accurate model was obtained when the task was split into a classification task to decide if the imaginary part of eigenvalues were stable (≤0) or not, and a regression model to calculate the eigenvalues once the classifier predicted the unstable class.


  • Open Access

    Enabling adaptive pedestals in predictive transport simulations using neural networks

    by A. Gillgren, E. Fransson, D. Yadykin, L. Frassinetti, P. Strand, Jet Contributors

    We present PEdestal Neural Network (PENN) as a machine learning model for tokamak pedestal predictions. Here, the model is trained using the EUROfusion JET pedestal database to predict the electron pedestal temperature and density from a set of global engineering and plasma parameters. Results show that PENN makes accurate predictions on the test set of the database, with R 2 = 0.93 for the temperature, and R 2 = 0.91 for the density. To demonstrate the applicability of the model, PENN is employed in the European transport simulator (ETS) to provide boundary conditions for the core of the plasma. In a case example in the ETS with varied neutral beam injection (NBI) power, results show that the model is consistent with previous studies regarding NBI power dependency on the pedestal. Additionally, we show how an uncertainty estimation method can be used to interpret the reliability of the predictions. Future work includes further analysis of how pedestal models, such as PENN, or other advanced deep learning models, can be more efficiently implemented in integrating modeling frameworks, and also how similar models may be generalized with respect to other tokamaks and future device scenarios.