ML-modelling the evolution of the quark-gluon plasma
Loading...
Date
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
В данной работе исследована применимость архитектуры U-net к решению
уравнений релятивистской гидродинамики, в частности, к предсказанию эволюции кварк-глюонной плазмы. Было разработано 3 эффективных нейросетевых подхода на основе архитектуры U-net, позволяющих построить эволюцию кварк-глюонной плазмы. На основе этих подходов была написана консольная утилита, способная строить эволюцию КГП вплоть до момента полного вымораживания. Применимость подходов была проверена на синтетических данных. Был произведен расчет азимутальных потоков.
In this paper, we have studied the ability of the U-net architecture to solve the equations of relativistic hydrodynamics, in particular, the ability to predict the evolution of the quark-gluon plasma. Three effective approaches based on the U-net architecture have been developed, which allow to construct the evolution of the quark-gluon plasma. Based on these approaches, a console application was implemented that is capable of building the QGP evolution up to the moment of complete freezeout. The applicability of the approaches was tested on synthetic data. The azimuthal flows were calculated.
In this paper, we have studied the ability of the U-net architecture to solve the equations of relativistic hydrodynamics, in particular, the ability to predict the evolution of the quark-gluon plasma. Three effective approaches based on the U-net architecture have been developed, which allow to construct the evolution of the quark-gluon plasma. Based on these approaches, a console application was implemented that is capable of building the QGP evolution up to the moment of complete freezeout. The applicability of the approaches was tested on synthetic data. The azimuthal flows were calculated.