mpc-horizon-tuning

Tune MPC horizon and cost matrices by selecting a prediction horizon N typically between 5 and 15 for R2R systems, and specifying state and control weights Q and R to balance tracking performance against actuator effort.

Design Q and R weights by prioritizing tension tracking in the Q matrix and actuator effort in the R matrix, while using an Algebraic Riccati Equation (ARE) to specify the terminal cost for stable control.

MPC horizon tuning affects disturbance rejection because a longer prediction horizon improves the controller's ability to anticipate and compensate for tension disturbances, though it increases the real-time computational load.

Yes, these MPC tuning guidelines are explicitly applicable to real-time tension-control scenarios in R2R systems operating with a sampling time of dt=0.01s, ensuring stable and efficient performance.

The terminal cost for MPC tension control is calculated via the Algebraic Riccati Equation (ARE), which provides a stabilizing terminal weight to ensure closed-loop stability within the finite prediction horizon.

Choose an MPC horizon length N typically between 5 and 15 for tension control, balancing the need for effective disturbance rejection against the available real-time computational capacity.