Argonne National Laboratory said scientists there developed an algorithm that could aid battery corrosion insight. The new technique accelerates the solving of material structures from patterns uncovered in X-ray experiments.
In a statement, the US national laboratory said it allows researchers to study certain properties, such as corrosion or battery charging and discharging, in real time.
The technique is called AutoPhaseNN, and is based on machine learning, which trains an algorithm on certain experimental data. It then chooses the most likely outcome of the current experiment.
The data used in this case are created by shining ultrabright X-ray beams from Argonne’s Advanced Photon Source (APS) on a material and capturing the light as they bounce off (diffraction).
Argonne said the APS is undergoing a massive upgrade, which will increase the brightness of its X-ray beams by up to 500 times. This means that more data will be gathered more quickly.
Yudong Yao, an Argonne X-ray physicist involved in the research, said: “With the kind of diffraction we’re doing, getting the phase information is a challenge; it’s like figuring out how all the pieces fit together solely based on the colours you can see on each piece.”
For conventional, supervised neural networks to solve this inverse problem, the researchers would have had to pair “broken puzzles” with fully assembled examples so that the neural network could have something to train against.
With an unsupervised neural network, the algorithm can learn to put together the puzzle from just the broken pieces. The resulting network is fast, accurate and, unlike conventional methods, capable of providing 3D images in real time to scientists.
