Machine Learning Applications for Process Control and Optimization in Steelmaking
Wednesday, 5 August 2020 • 10–11 a.m. EDT
Giovanni Bavestrelli; Tenova S.p.A.
As part of Tenova’s digital transformation, the digital team at Tenova has worked closely with various business units to identify opportunities for application of artificial intelligence (AI) techniques where they could bring value to customers in the steel industry. This paper discusses some of these initiatives applied to the control and optimization in steel production plants:
Application of machine learning (ML) to data from a basic oxygen furnace.
Application of ML to control cassette penetration and strip tension in a tension leveler machine.
Application of convolutional neural networks for automatic classification of scrap material in a steel plant.
Application of ML to predict metallization and carbon content in a direct reduction plant.
Model to predict NOx emissions in a walking beam furnace.
The paper describes the various machine-learning models developed, the benefits obtained, as well as lessons learned during the process.
Chris Burnett; Thermo Fisher Scientific
IST’s Electrical Applications/Sensors Subcommittee and Digitalization Applications Technology Committees.