Tyumen Researcher Finds Way to Speed Up Refinery Operations
To do so, he proposed combining the physical models currently used for calculations with elements of artificial intelligence.
Kirill Motaev, a young researcher from Tyumen State University, presented a technology at an international conference at ITMO University in St. Petersburg that can significantly accelerate refinery operations. It combines the accuracy of the physical models engineers work with and the speed of artificial intelligence. The new hybrid performs calculations a thousand times faster than conventional tools, and it runs on an ordinary personal computer.
The developers managed to embed AI elements into the traditional model. This allowed them to combine their strengths and offset their weaknesses.
"In practice, this means the following: refinery engineers will be able to select optimal conditions — such as temperature, pressure, and feedstock composition — for producing the highest-quality gasoline literally within hours rather than months," Motaev said.
