Rules-based systems for improved monitoring and guidance of reheating furnaces

General Modem reheating furnaces in rolling mills have to supply a product with the target, homogeneous temperature and low scale formation, while ensuring the minimum energy consumption and as maximum as possible productivity. The management of the continuous reheating of slabs having varying size, steel grade and discharging temperature is an extremely complex task. Also the most skilful operator may be no longer capable of making the required fine adjustments completely manually. Hot strip mill reheating furnaces are nowadays regularly equipped with process control systems based on mathematical models containing the known physical constraints of the process and its environment. Main aim of these process control systems is to ensure optimal reheating conditions for the different steel qualities, maximising plant productivity and minimising fuel consumption. Nevertheless, these control systems are only as good as the mathematical model of the furnace and the reliability of measurements and final control elements. Unknown operational conditions can not be controlled efficiently by such a system. For this reason in reality the furnaces are often controlled manually by experienced operational staff. In some cases only 60% of operational time is controlled by the furnace control system. The remaining 40% are based on new operational conditions and therefore controlled manually by the operational staff. The utilisation of the operator experience and knowledge within a computer based control system might drastically increase the reheating process efficiency and productivity and the operational staff should be relieved. Objectives The overall objective of the project is the development and implementation of intelligent monitoring and guidance of reheating furnaces for optimised operation, based on artificial intelligence techniques, in terms of: • component and process on-line diagnosis (MEFOS, CSM); • estimation of the temperature and furnace atmosphere set points (BFI). The systems are able not only to prevent or to diagnose malfunctions in the furnace, but also to monitor the re-heating process according to different production ranges, providing the operator with guidance when deviating from correct behaviour. The innovative idea of the present project is to extend the conventional process control techniques by combining the results of on-line physical models calculations with expert system rules and statistical methods. The ability to reconcile significant amount of data from a number of disparate sources, continuously apply expert rules and statistical methods as a form of analysis and decision support to operators, provide the basis of improved decision making and control.