Russian scientists published datasets for correction of Coriolis flowmeter readings

Scientists at South Ural State University have proposed a new way to work on the correction of Coriolis flowmeter errors. SUSU specialists, together with colleagues from Oxford, developed a new machine-learning method based on a data set. An article on a reference data set proposed for comparing correction methods was published in one of the most ranked journals of the first quartile "Flow Measurement and Instrumentation".

Dataset for error-free Coriolis flowmeters

Coriolis flowmeters have greatly facilitated the work of calculating the mass flow rate, density, and temperature of liquids and gases. Instrument readings are accurate in a single-phase medium, but in a two-phase one, for example, when air bubbles appear in water, the measurement accuracy drops sharply. Therefore, researchers are developing ways to correct errors using machine learning methods.

SUSU scientists have noticed that all these methods are not universal. Scientists made a series of experiments in various conditions using different kinds of flow meters. Therefore, it is incorrect to compare the results with each other. Representatives of the School of Electronic Engineering and Computer Science of SUSU found the opportunity to fix this.

The head of the International Laboratory of Technical Self-Diagnostics and Self-Monitoring of Instruments and Systems at SUSU, Professor of Oxford University (Great Britain) Manus Henry, in the TUV-NEL laboratory, collected a dataset with measurements of Coriolis flowmeters. They have been published for public use.

"The main goal of our work is to offer a data set that is open and common for everyone, on which it would be possible to compare the various methods developed for correcting errors of Coriolis flowmeters. In machine learning, there are many similar datasets on which scientists "hone" their methods and compare the results correctly. However, there were still no datasets with all Coriolis flowmeter measurement data open, and our work fills this gap, "Olga Ibryaeva, associate professor at the System Programming Department of the School of Electronic Engineering and Computer Science, explained.

The published data was the first to be used by Olga Ibryaeva and Vyacheslav Barabanov, a graduate student at the School of Electronic Engineering and Computer Science. They selected the architecture of the neural network, where the error of measurements by Coriolis flowmeters is less than 1%.

A valuable resource for scientists created by students

The authors of the article believe that the datasets obtained will be used to develop other methods correcting the measurement of the mass and density of the liquid. Therefore, all data sets obtained in a two-phase medium were published for public access on the website of SUSU. It was created by a second-year bachelor of the School of Electronic Engineering and Computer Science, Denis Lebedev, and Alexander Zhulev. Now they are working on improving the site and a forum where scientists could communicate.

SUSU professors hope that the site, will become a recognizable source of reference data for developing methods in the field of correction of Coriolis flowmeter measurements, as well as research results.

"We continue to develop other methods that may give an even better result on this dataset, and encourage scientists to do so. Our site also has other datasets obtained in a two-phase environment, and in the future, we plan to supplement them with data obtained when the flow meter is operating in a three-phase environment. We also have the possibility of loading our datasets with other researchers," Olga Ibryaeva said.

South Ural State University is a university of digital transformations where innovative research is being held on most of the priority areas for science and technology development. In accordance with the strategies of scientific and technological development of the Russian Federation, the university is focused on the development of large scientific interdisciplinary projects in the field of IT, materials science, and ecology. In the framework of these areas, metallurgy, engineering, energy, urban infrastructure, and facilities are investigated.

SUSU is a participant in the 5-100 Project, designed to increase the competitiveness of Russian universities among the world's leading research and academic centers.

More information:
www.sciencedirect.com/science/ … 0955598620300698?via%3Dihub

Provided by South Ural State University