Model predictive control for constant setpoint tracking

Paper Title:

Model Predictive Control for setpoint tracking

Published on:

5 March 2024

Primary Category:

Optimization and Control

Paper Authors:

Daniel Limon,

Antonio Ferramosca,

Ignacio Alvarado,

Teodoro Alamo

Bullets

Key Details

•

MPCT adds artificial steady state as optimization variable

•

Relaxes terminal constraint using artificial state

•

Penalizes deviation of artificial state from setpoint

•

Ensures feasibility and stability for changing setpoints

•

Achieves larger region of asymptotic stability than standard MPC

Explore the topics in this paper

artificial steady state

asymptotic stability

model predictive control

optimal control

tracking control

AI generated summary

Model predictive control for constant setpoint tracking

This paper presents a model predictive control (MPC) approach called MPC for tracking (MPCT) to steer a constrained linear system to track constant setpoints. It ensures recursive feasibility and asymptotic stability even when setpoints change. The MPCT adds an artificial steady state as a decision variable, relaxes the terminal constraint, and penalizes deviation from the artificial state to the true setpoint. This allows it to recursively re-optimize to track new setpoints. Under mild assumptions, the region of asymptotic stability and domain of attraction are larger than standard MPCs, and local optimality can be achieved.