Inferential control for rolling mills

A robust inferential controller is developed for hot strip rolling mill ten sion control. Tension cannot be measured, therefore a nonlinear model is us ed to infer the tension from the looper angle. The model uses constant phys ical parameters while the actual physical system will be uncertain, therefo re quantitative feedback theory is used to design a tension controller, whi ch will provide robust performance despite uncertainty from both the physic al system and the inference model or soft sensor.