Purpose and goal
The AOD process (Argon Oxygen Decarburization) contributes to the production of steel with high purity and specific properties, which is essential to achieve the high quality required for stainless steel. As the valves in the AOD converter age, it can give rise to undesirable behavior that can lead to to quality defects or cancellation.
The objective of the project is to use machine learning and data from control systems to detect abnormal behavior of valves for AOD converters in real time.
Expected effects and result
Two case studies at two independent factories shall result in two demonstrators. In addition, the project will result in a guide for AI implementations in existing production environment in the steel industry.
Planned approach and implementation
Control systems used for regulation and monitoring contain large amounts of data. With machine learning algorithms, normal behavior as well as abnormal behavior can be detected. In the proposed project, algorithm development takes place based on knowledge of the process. During a learning period (training period), the competence of experienced expertise is used to identify recurring errors. Then a validation phase takes place and finally the valve monitoring takes place automatically.
The project description has been provided by the project members themselves and the text has not been looked at by our editors.
