We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.SY

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Systems and Control

Title: A Semi-Definite Programming Approach to Robust Adaptive MPC under State Dependent Uncertainty

Abstract: We propose an Adaptive MPC framework for uncertain linear systems to achieve robust satisfaction of state and input constraints. The uncertainty in the system is assumed additive, state dependent, and globally Lipschitz with a known Lipschitz constant. We use a non-parametric technique for online identification of the system uncertainty by approximating its graph via envelopes defined by quadratic constraints. At any given time, by solving a set of convex optimization problems, the MPC controller guarantees robust constraint satisfaction for the closed loop system for all possible values of system uncertainty modeled by the envelope. The uncertainty envelope is refined with data using Set Membership Methods. We highlight the efficacy of the proposed framework via a detailed numerical example.
Comments: Accepted for European Control Conference (ECC), May 2020, Saint Petersburg, Russia
Subjects: Systems and Control (eess.SY); Optimization and Control (math.OC)
Cite as: arXiv:1910.04378 [eess.SY]
  (or arXiv:1910.04378v2 [eess.SY] for this version)

Submission history

From: Monimoy Bujarbaruah [view email]
[v1] Thu, 10 Oct 2019 06:07:06 GMT (269kb,D)
[v2] Sat, 15 Feb 2020 23:36:23 GMT (286kb,D)

Link back to: arXiv, form interface, contact.