Faster fusion reactor calculations thanks to machine learning


plasma

Credit: CC0 Public Domain

Fusion reactor technologies are well positioned to contribute to our future energy needs in a safe and sustainable way. Numerical models can provide researchers with information on the behavior of the fusion plasma, as well as valuable information on the efficiency of the reactor design and operation. However, modeling the large number of plasma interactions requires several specialized models that are not fast enough to provide data on the design and operation of the reactor. Aaron Ho of the Nuclear Fusion Science and Technology group in the Department of Applied Physics has explored the use of machine learning approaches to accelerate the numerical simulation of turbulent transport of central plasma. Ho defended his thesis on March 17.

The ultimate goal of fusion reactor research is to achieve a net energy gain in an economically viable manner. To achieve this goal, large and intricate devices have been built, but as these devices become more complex, it becomes increasingly important to take a predictive approach first regarding their operation. This reduces operational inefficiencies and protects the device from serious damage.

To simulate such a system requires models that can capture all the relevant phenomena in a fusion device, that are accurate enough that the predictions can be used to make reliable design decisions, and that are fast enough to quickly find viable solutions.

Neural network-based model

For his Ph.D. In the research, Aaron Ho developed a model to satisfy these criteria using a model based on neural networks. This technique allows a model to retain both speed and accuracy at the cost of collecting data. The numerical approach was applied to a reduced order turbulence model, QuaLiKiz, which predicts the amounts of plasma transport caused by microturbulence. This particular phenomenon is the dominant transport mechanism in tokamak plasma devices. Unfortunately, your calculation is also the limiting speed factor in today’s tokamak plasma modeling.

Ho successfully trained a neural network model with QuaLiKiz evaluations while using experimental data as training input. The resulting neural network was then coupled to a larger integrated modeling framework, JINTRAC, to simulate the core of the plasma device.

Simulation time was reduced from 217 hours to just two hours.

Neural network performance was evaluated by replacing the original QuaLiKiz model with Ho’s neural network model and comparing the results. Compared to the original QuaLiKiz model, Ho’s model considered additional physical models, duplicated the results with 10% accuracy, and reduced simulation time from 217 hours on 16 cores to two hours on a single core.

Then, to test the effectiveness of the model outside of the training data, the model was used in an optimization exercise using the coupled system in a plasma acceleration scenario as a proof of principle. This study provided a deeper understanding of the physics behind the experimental observations and highlighted the benefit of fast, accurate, and detailed plasma models.

Finally, Ho suggests that the model can be extended for other applications, such as controllers or experimental design. He also recommends extending the technique to other physical models, as it was observed that turbulent transport predictions are no longer the limiting factor. This would further enhance the applicability of the embedded model in iterative applications and enable the validation efforts required to move its capabilities closer to a truly predictive model.


New turbulent transport model shows multiscale fluctuations in heated plasma


Provided by Eindhoven University of Technology

Citation: Faster Fusion Reactor Calculations Thanks to Machine Learning (2021, March 22) Retrieved March 23, 2021 from https://phys.org/news/2021-03-faster-fusion-reactor-machine.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or investigation, no part may be reproduced without written permission. The content is provided for informational purposes only.



Source link