Robust decoupling MPC for linear systems with bounded disturbances.

Area of Science:

• Control Systems Engineering
• Automation and Robotics
• Applied Mathematics

Background:

• Constrained linear systems are susceptible to bounded disturbances, impacting performance and stability.
• Traditional decoupling methods may struggle with recursive feasibility and robust constraint satisfaction.
• Accurate tracking of piecewise constant references in the presence of uncertainties remains a challenge.

Purpose of the Study:

• To develop a robust decoupling Model Predictive Controller (MPC) for constrained linear systems with bounded disturbances.
• To ensure recursive feasibility, robust constraint satisfaction, and minimize cross-coupling during set-point changes.
• To achieve offset-free tracking of piecewise constant references.

Main Methods:

• Integration of explicit decouplers with an augmented state-space representation.
• Application of a robust MPC algorithm tailored for piecewise constant reference tracking.
• Utilizing an artificial target based on nominal prediction for offset-free tracking.

Main Results:

• Demonstrated recursive feasibility and robust constraint satisfaction under bounded disturbances.
• Significant reduction in cross-coupling effects during set-point transitions.
• Effective offset-free tracking of piecewise constant references validated through case studies.

Conclusions:

• The proposed robust decoupling MPC strategy effectively addresses constraints and disturbances in linear systems.
• The approach offers flexibility in choosing decouplers and robust MPC algorithms.
• The method provides a reliable solution for set-point changes and reference tracking applications.