PRESETTING OF ROLLING STAND PARAMETERS BY GENETIC ALGORITHMS

摘要

Bland-Ford mathematical model of rolling process is often used in metallurgy. The model allows to analyze rolling process according to data measured on the mill stand to get immeasurable variables for rolling control. The values obtained by modeling a currently rolled coil are used for mill presetting for the next coil. However, to achieve real-time rolling control, the Bland-Ford model is modified applying an artificial intelligence approach instead of analytical solution of a set of nonlinear differential equations. The genetic algorithm with hierarchical structure was used instead of classical iterative method.