Development of Advanced Furnace Control Algorithm

In the continuous work on improving product quality and reducing energy consumption at The Danish Steel Works Ltd. the importance of optimal furnace operation has been recognized in many di erent contexts. Therefore a joint project with The Department of Automatic Control at Lund Institute of Technology has been started with the purpose of developing a novel slab temperature control algorithm for the reheat furnace control systems. The article describes the results of this work so far by giving a short description of the ideas behind the new algorithm and presenting simulations where the performance of the new algorithm is compared to the performance of a conventional PI-controller. The model for the simulations are found from data collected from one of the reheat furnaces on The Danish Steel Works Ltd. The simulations show that the new algorithm is able to reduce the standard deviation from the desired nal temperature with more than 15% compared to the temperature deviation obtained with the PI-control algorithm.