Futurists of the past used to imagine that breakthroughs in technology and science could create a utopian world fueled by boundless clean energy. Now, an artificial intelligence model from researchers at Princeton may have proven them right. Or at least, it’s gotten us a step closer.
Fusion – the nuclear reaction in which two or more atomic nuclei combine and form new nuclei and subatomic particles – has long been the dream as an energy source: it’s non-polluting, safe, and virtually limitless, producing nearly four million times as much energy by mass as burning fossil fuels.
Unfortunately, there’s a problem. Fusion is really, really hard to do: it requires the kinds of temperatures and pressures that are found in the hearts of stars. Since we can’t really get those exact conditions in a lab on Earth, the relatively few examples of human-created fusion have relied on a workaround: normal terrestrial pressure, and temperatures more than ten times that of the Sun’s core.
At those temperatures, the fuel needed for the reaction can’t exist in a solid or liquid state, and it’s not even in there as a gas – it’s plasma. Therein lies another problem: this state of matter is so energetic and superheated that it’s easy for the fuel to “tear” – to lose stability and escape the magnetic fields keeping it within the reactor – thus putting an end to any fusion within milliseconds.
It’s precisely this problem that the Princeton team claims to have solved.
“Previous studies have generally focused on either suppressing or mitigating the effects of these tearing instabilities after they occur in the plasma,” explained first author of the new paper Jaemin Seo, now an assistant professor of physics at Chung-Ang University in South Korea, in a statement. “But our approach allows us to predict and avoid those instabilities before they ever appear.”
Their answer: an artificial intelligence (AI) trained on previous experiments at the DIII-D National Fusion Facility in San Diego.
“By learning from past experiments, rather than incorporating information from physics-based models, the AI could develop a final control policy that supported a stable, high-powered plasma regime in real time, at a real reactor,” said research leader Egemen Kolemen, associate professor of mechanical and aerospace engineering and the Andlinger Center for Energy and the Environment and research physicist at the Princeton Plasma Physics Laboratory (PPPL).
Like any AI model, it doesn’t really understand what it’s doing on a deep level – but it doesn’t need to. The team fed the program data about real-time plasma characteristics from previous experiments and set it the challenge of predicting – and, crucially, avoiding – tearing instabilities.
“We don’t teach the reinforcement learning model all of the complex physics of a fusion reaction,” explained Azarakhsh Jalalvand, a research scholar in Kolemen’s lab and coauthor of the paper. “We tell it what the goal is – to maintain a high-powered reaction – what to avoid – a tearing mode instability – and the knobs it can turn to achieve those outcomes. Over time, it learns the optimal pathway for achieving the goal of high power while avoiding the punishment of an instability.”
After myriad simulations, which were able to be tweaked and refined by human observers, the team tried the AI out for real at the D-III D facility. The model proved itself capable of predicting tearing instabilities up to 300 milliseconds in advance – not much to a human, but plenty of time for the AI to act, changing parameters such as the shape of the plasma or the strength of the beams inputting power to the reaction in order to keep the plasma stable.
So is unlimited clean energy just around the corner? Not quite. Plasma instability is far from the only problem with fusion – and tearing is only one type of possible plasma instability.
But what the paper does show, the team says, is a pretty good proof of concept: “We have strong evidence that the controller works quite well at DIII-D, but we need more data to show that it can work in a number of different situations,” Seo said. “We want to work toward something more universal.”
The paper is published in the journal Nature.
Source Link: AI Just Cleared A Big Hurdle On The Road To Nuclear Fusion Energy
