The increasing demands on heating and heat treatment furnaces in terms of temperature distribution in slabs or strips, optimal use of energy for reaching the target temperatures as well as many other factors require modern control strategies with the aim to ensure the best possible system efficiency including the up- and downstream processes. The control and automation supports the workers at the plant and provides a higher product quality. In the past, as from the historical view, there were often simple and robust control solutions for each parameter of the furnace. One step further to better control solutions for furnaces is to build up a non-linear model with predictive control (MPC). Such models are used to represent the behaviour of very complex dynamical systems. But the additional complexity of the MPC control algorithm is not always generally needed to provide adequate control, which is often controlled well by generic PID controllers. Common dynamic characteristics that have an adverse effect on PID controllers include large time delays and high-order dynamics. Within these models there is a prediction of the change in the dependent variables of the modelled system that will be caused by changes in the independent variables. In the future it might be possible that control problems will get more and more complex. The background to this is the strong customer orientation of under the conditions of a highly flexible production as it is at the end of the value chain, that the demand for accuracy for high-quality steels is highest. The required automation and technology in this field is the introduction of methods of self-optimization, self-configuration, self-diagnosis, cognition and intelligent support of workers in their increasingly complex work.

1. http://www.dotxcontrol.com/en/downloads/software.html.