The green energy revolution promised by nuclear fusion is now a step closer, thanks to the first successful use of a cutting-edge artificial intelligence system to shape the superheated hydrogen.
The use of artificial intelligence could be a breakthrough in the search for electricity generated from nuclear fusion.
Federico Felici, a physicist at the Swiss Federal Institute of Technology in Lausanne, believes that Artificial Intelligence will play a big role in the future control of tokamaks.
What is the difference between fusion and fissure?
Felici is a lead author of a new study. He said future experiments at the TCV in Lausanne will look for ways to integrate artificial intelligence into the control of fusion reactor.
The artificial intelligence system on the TCV was tested by Felici and his colleagues at the EPFL's Swiss Plasma Center.
The doughnut-shaped fusion reactor is the type that seems most promising for controlling nuclear fusion; a tokamak design is being used for the massive international ITER project being built in France.
Felici explained that the tokamak is controlled by 19 magnetic coils that can be used to shape and position hydrogen inside the fusion chamber.
A set of independent computerized controllers governs the coils, one for each aspect of the plasma that features in an experiment, that are programmed according to complex control engineering calculations, depending on the particular conditions being tested. He said that the new system was able to manipulate the plasma with a single controller.
The deep reinforcement learning system developed by DeepMind was first trained on simulations of the tokamak, a cheaper and much safer alternative to the real thing.
It takes several hours for the computer simulations to mimic real-time tokamak operation. The TCV's experimental condition can change from day to day, and so the developers needed to take those changes into account in the simulations.
The actual tokamak was coupled to the artificial intelligence when the training process was complete.
For a maximum of 3 seconds, the TCV can sustain a superheated hydrogen plasma. Between 30 and 35 such shots are done each day, and it takes 15 minutes to cool down and reset.
Over the course of several days, a total of about 100 shots were done with the TCV, he said.
Science fact or fiction? There is a plausibility to 10 sci-fi concepts.
He said that the new ways of shaping the tokamak resulted in greater control of the entire fusion process.
The so-called snowflake shape, which is thought to be the most efficient configuration for fusion, was one of the most common configurations that the artificial intelligence was able to position and shape.
He said that it was possible to create separate upper and lower rings of plasma within the chamber, which had never been attempted before.
Felici said that it was easy to create the droplet shape with the machine learning.
He said that the researchers saw that the artificial intelligence was using the magnetic coil to control the plasmas inside the chamber in a different way than would have happened with the standard control system.
He said that we can apply the same concepts to more advanced problems.
The ITER project is a massive tokamak that is projected to achieve full-scale fusion in about 25 years. Proponents hope ITER will pioneer new ways of using nuclear fusion to generate usable electricity without carbon emissions and with low levels of radioactivity.
The TCV experiments will inform designs for the DEMO fusion reactor, which are seen as successors to ITER that will supply electricity to power grids, something that ITER is not designed to do. Europe's EUROfusion reactor is projected to begin operations in 2051, and several countries are working on designs for the DEMO reactor.
It was originally published on Live Science.
All Rights Reserved © 2024
