Computation of Rolling Stand Parameters by Genetic Algorithm.docVIP

Acta Polytechnica Hungarica Vol. 5, No. 2, 2008 Computation of Rolling Stand Parameters by Genetic Algorithm Franti?ek ?urovsky, Ladislav Zboray, ?elmíra Ferková Department of Electrical Drives and Mechatronics Technical University of Ko?ice Letná 9, 042 00 Ko?ice, Slovakia e-mail: frantisek.durovsky@tuke.sk Abstract: Mathematical model of rolling process is used at cold mill rolling on tandem mills in metallurgy. The model goal is to analyse rolling process according to process data measured on the mill and get immeasurable variables necessary for rolling control and optimal mill pre-set for next rolled coil. The values obtained by model are used as references for superimposed technology controllers (thickness, speed, tension, etc.) as well. Considering wide steel strip assortment (different initial and final thickness, different hardness), and fluctuation of tandem mill parameters (change of friction coefficient, work rolls abrasion, temperature fluctuation, etc.) the exact analysis of tandem is complicated. The paper deals with an identification of friction coefficient on a single rolling mill stand by a genetic algorithm. Mathematical description of tandem mill stand is based on the modified Bland-Ford model. Results are presented in graphical form. Keywords: cold rolling, Bland-Ford model, mathematical model, genetic algorithm 1 Introduction Cold rolling of steel is very complex process. Knowledge of conditions in the roll bite is essential to achieve a good quality of final production. Some of its parameters may be determined exactly by measurement on the mill stand (geometrical dimensions, rolling force, front and back tensions, roll velocity), another only approximately by a suitable mathematical model (e.g. hardening of the processed material, friction coefficient between rolls and strip, strip temperature and flatness)