A new tractable tube-based robust data-driven MPC for unknown discrete-time LTI systems using one noisy trajectory, simplex-constrained Hankel coefficients, and certified RPI sets to ensure recursive feasibility and practical ISS via a convex QP.
Title resolution pending
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Necessary and sufficient conditions are given for contraction of affine input-output systems, implementability of contractive references, and their realization via linear or affine feedback.
Reference condensation compresses preview trajectories into a single setpoint for MPC via least-squares projection, keeping parameter dimension independent of horizon length.
In linear systems with parametric uncertainty and Gaussian noise, MPC policy dependence on posterior covariance peaks at high uncertainty and vanishes as it contracts, with the dual controller outperforming certainty-equivalent MPC on both regulation performance and model accuracy.
A data-driven procedure constructs robust positively invariant tube sets from noisy data of unknown LTI systems and certifies them for use in tube-based robust predictive control.
SQP and LPV-MPC are shown to be equivalent under a specific scheduling choice and FTC embedding, enabling a zero-order LPV-MPC approach for faster real-time control.
citing papers explorer
-
Stability, Contraction, and Controllers for Affine Systems
Necessary and sufficient conditions are given for contraction of affine input-output systems, implementability of contractive references, and their realization via linear or affine feedback.
-
Reference Condensation for Model Predictive Control with Preview
Reference condensation compresses preview trajectories into a single setpoint for MPC via least-squares projection, keeping parameter dimension independent of horizon length.
-
The Separation Principle and the Dual-Certainty Equivalence Gap in Model Predictive Control
In linear systems with parametric uncertainty and Gaussian noise, MPC policy dependence on posterior covariance peaks at high uncertainty and vanishes as it contracts, with the dual controller outperforming certainty-equivalent MPC on both regulation performance and model accuracy.