|GA-Based Optimisation of a Continuous Walking Beam Reheating Furnace|
|Jonathan Broughton - University|
Mahdi Mahfouf - University of Sheffield
Derek Linkens - University of Sheffield
The paper focuses on the optimisation of schedules for a walking beam reheat furnace at the Corus Thrybergh Bar Mill. In this research a novel approach was developed for the assessment and refinement of these schedules. The model-based furnace control was reverse engineered to form a discrete-event simulator to assess the performance of schedules and an in-house multi-objective genetic algorithm was developed as the schedule optimiser. Fuel consumption and heating quality measures were incorporated in the optimisation process. The system measures the performance of a schedule taking into account the performance of the control to process that schedule, thereby tightening the link between scheduling and control.