Beschreibung
Almost all flat steel production involves processing in a tandem hot strip finishing mill, where the mass flow significantly impacts product quality and process stability. Mass flow can be categorized as longitudinal or lateral based on its direction. Uniform longitudinal mass flow ensures constant strip tension, improving thickness quality and preventing necking. A measurement system for mass flow is described, allowing validation and tuning of slip models. An adaptation approach using an additional disturbance input improves strip velocity accuracy compared to traditional models like Sims'. In mass flow control, roll speeds are synchronized in a cascade master-slave structure, with the downstream mill stand as the master. Since slip between rolls and strip varies, an adaptive synchronization controller is proposed, requiring minimal computation and integrating seamlessly into existing controls. A robust synchronization law addresses sensor faults, switching to a fault-tolerant form when needed. Stability is proven using Lyapunov's method, and simulations validate the system. Lateral mass flow affects the strip's lateral movement, which, if uncontrolled, can damage the strip or plant. An extended lateral mass flow model supports a model predictive controller (MPC) that respects variable limits and ensures stability. The model, validated with industrial data, integrates seamlessly into existing plant controls. A dynamic observer estimates unmeasurable strip angles and positions in the roll gap, offering soft-sensed camber measurements and valuable operator insights. Simulations using a validated mathematical model confirm the effectiveness of the proposed lateral control strategy.
