제어로봇시스템학회:학술대회논문집
- 제어로봇시스템학회 2003년도 ICCAS
- /
- Pages.723-727
- /
- 2003
A study on the improvement of thickness accuracy in a plate mill
- Lee, Duk-Man (Rolling Technology and Process Control Group, Technical Research Center, POSCO)
- 발행 : 2003.10.22
초록
In this paper, two methods are discussed for good rolling force prediction in a plate mill. One is about the development of a long and a short learning scheme using a Neural Network for normal rolling and the other is about a mathematical model improvement by considering microstructural changes for controlled rolling. The research results are implemented in a on-line system of Pohang Works in POSCO and the field tests have showed that the prediction accuracies of rolling force are highly improved.