Help Desk -
9958816305, 9810335381

Severstal Regulates Slab Rolling Using Machine Learning Algorithms

Steel News - Published on Thu, 14 Nov 2019

Image Source:
The sheet rolling shop for the production of tube steel at the Cherepovets Steel Mill together with Severstal Digital implemented a solution based on machine learning models that automatically adjusts the speed of slab rolling at 5000 mill. Previously, specialists in the sheet rolling workshop controlled the speed of rolling slabs manually, adjusting the speed with a rheostat. At the same time, the mill operator had to simultaneously control the unit, as well as monitor the passage and quality of the metal in order to avoid defects. When determining the rolling speed, the product mix and rolling cycles were not always taken into account, so the indicators could be underestimated and the mill did not work at full capacity.

To increase the productivity of mill 5000, a solution was developed based on machine learning algorithms. Speed control is carried out taking into account the rental mode formed in the metal tracking system, sheet length and width, roll gap, steel grade, temperature and other parameters. The model comprehensively evaluates these indicators online and sets the optimal rental speed in a given cycle. The metal condition is monitored regularly, so the parameters in each metal pass are adjusted. As a result, the maximum rolling speed indicator increases: from 3.2 m / s set by the operator to 4.5 m / s set by the machine. As a result, the number of rolled slabs is also increasing. The expected increase in mill productivity will be from 6% to 9%, depending on the rolled product mix. Besides,

Source :

Posted By : Rabi Wangkhem on Thu, 14 Nov 2019
Related News from Steel segment