Shulou(Shulou.com)05/31 Report--

Artificial neural networks-algorithms inspired by brain connections-have "learned" to perform tasks ranging from pedestrian detection in self-driving cars to medical image analysis to language translation. Now, researchers at the University of California, San Diego are training artificial neural networks to predict new stable materials.

"predicting the stability of materials is a core issue in materials science, physics and chemistry," said Xiugongpin, a professor of nanoengineering and senior author at the Jacobs School of Engineering at the University of California, San Diego. " On the one hand, you have traditional chemical intuition, such as LinusPauling's five laws, which describe the stability of crystals in terms of radius and ion accumulation. On the other hand, you have expensive quantum mechanical calculations to calculate the energy obtained from crystal formation, which must be done on a supercomputer. What we do is connect the two worlds with an artificial neural network.'

Training the artificial neural network to predict the crystal formation energy requires only two inputs-to make up the electronegativity of the atom and the ion radius of 3354. Ong of the Materials Virtual Lab and his team developed a model that can identify stable materials in two types of crystals, called garnet and perovskite. These models are 10 times more accurate than previous machine learning models, and they are fast enough to effectively screen out thousands of materials in a matter of hours. This work is described in detail in a paper published on September 18 of the year Nature Communications.

Garnet and perovskite are used in LED lamps, rechargeable lithium-ion batteries and solar cells. These neural networks have the potential to greatly accelerate the discovery of new materials for these and other important applications. Students of the anger Materials Virtual Lab.

The team has made their model available to the public through http://crystals.ai 's online app, which allows others to use these neural networks to calculate the formation energy of any garnet or perovskite component in flight.

The researchers plan to extend the application of neural networks to other crystal prototypes and other material properties.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.