In-time rating of continuously cast semis by means of hybrid quality models
The aim of this project was the application of modern techniques to develop models for the quality estimation of cast semis. These models should be applicable online to the process control systems in order to support operators in process monitoring and control and thus to improve the efficiency on the steel plants. During the project at three different casting plants, the occurrence of quality defects at slabs and billets was examined. Data describing the production conditions and features of the raw material were collected to build databases representing the production of cast semis. From inspection stands or by individual analysis of slabs and billets, information was collected about the quality state of the cast semis or the final products. For example, slabs were inspected especially with regard to depressions; billets and rolled bars and wires were checked concerning the occurrence of surface defects like shells. During the project, tools have been developed to visualise online complex information from the casting process for operators, to analyse the dependencies of process and quality data and to include operators� and experts� experiences into rule-based classification modules. Different methods have been tested to monitor and identify process conditions during casting and to estimate the quality of semis. The transferability of methods was proven by the application to different casters, and developed models were implemented at the caster control systems to recognise quality-influencing process conditions and to support operators by suitable presentation of model results.
Bibliographic Reference: EUR 21984 EN (2006), 94pp. Euro: 25
Availability: Katalogue Number: KI-NA-21984-EN-S The paper version can be ordered online and the PDF version downloaded at: http://bookshop.europa.eu
ISBN: ISBN: 92-79-01097-2
Record Number: 200618561 / Last updated on: 2006-04-28
Original language: en
Available languages: en