MOSA, the MODEL OPTIMIZED for STEEL ANNEALING
Optimize your line!
Today, steel producers face a high variability of products on their schedule as well as the now coming new steel generations. The pressure from the market for high quality products at the best possible price has never been so high.
On-line controlling heavy inertial equipment like continuous annealing furnaces requires skilled operators. In response to this situation, Drever International implements MOSA, innovative developments in mathematical models for steel annealing control.
The concept is based on a flexible multi-purpose kernel application designed for all kinds of lines and their products:
- Carbon steel annealing, galvanizing, galvannealing
- Stainless austenitic and ferritic annealing
- Silicon steel annealing
- Bright annealing, etc.
These histogram graphs of strip temperature deviation covering steady states and transient states illustrate the continuous improvement over the last decade. The left graph is from a line driven by operators. The right graphs show results under math model operation over a certain period of time. Thanks to accurate self-learning technology high model accuracy is reached.
The most demanding process sets the standards of the MOSA kernel, which will be applied on all kinds of lines. Here, the results come from the production of automotive quality steel, in a plant with high variability of the products. As usual on high performance lines, the furnace and the line speed are fully under mathematical model control. The Peak Metal Temperature (pale red) shows a complex behaviour, which is part of the advanced transition management concept. The PMT is strictly maintained within its quality limits (fuchsia).
Quality is a key factor but the respect of the quality at the highest productivity as shown on the previous figure is a double challenge.
The ‘Productivity’ figure shows data from a line originally designed for 400,000 tons/year. The line has been pushed to 500,000 tons/y and lately the model has been revamped with a new generation MOSA. On this heavily loaded line, the productivity could be even more increased thanks to current mathematical model technology.
