This Machine Learning Algorithm Could Improve Lithium-Ion, Fuel Cell Performance

The algorithm will enable researchers to visualize and explore battery microstructures virtually through 3D simulations


Researchers at Imperial College London claim to have developed a new machine-learning algorithm that could improve the design and performance of lithium-ion batteries and fuel cells.

They claimed that the algorithm would also allow researchers to understand the microstructure of fuel cells better. This will help run simulations that could enable them to improve the performance of these battery technologies, they noted.

Lithium-ion batteries are the most commonly used in electronic devices and vehicles. A fuel cell, on the other hand, is a device that converts chemical energy into electrical energy using oxidizing agents through an oxidation-reduction (redox) reaction. Improvements in these technologies would have widespread implications across several product sectors.

In their paper, the researchers explained that the performance of fuel cells is dependent on their microstructure and how the pores inside their electrodes are shaped and arranged. This determines how much power fuel cells can generate and the speed at which they can be charged and discharged.

The research paper published in the npj Computational Materials journal said that the algorithm could help reduce the volume of electrochemical simulations required to test the performance of a particular microstructure design during optimization.

The technique called the “deep convolutional generative adversarial networks” (DC-GANs),” will enable researchers to visualize and explore these pores virtually through three-dimensional (3D) simulations.

“Our technique helps us zoom right in on batteries and cells to see which properties affect overall performance. Developing image-based machine learning techniques like this could unlock new ways of analyzing images at this scale,” said Andrea Gayon-Lombardo, lead author of the paper from Imperial’s Department of Earth Science and Engineering

Previously, Japanese researchers said they have developed a new electrode material that they claim will make lithium batteries cheaper, more stable, and capable of holding more charge for longer periods.

Earlier, Researchers at Penn State University claimed to have developed a lithium-ion battery that is safe and has power and can last up to one million miles. A team of researchers at the Penn State’s Battery and Energy Storage Technology (BEST) Center developed the battery.