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: Model predictive control design for dynamical systems learned by Long Short-Term Memory Networks

Abstract: This paper analyzes the stability-related properties of Long Short-Term Memory (LSTM) networks and investigates their use as the model of the plant in the design of Model Predictive Controllers (MPC). First, sufficient conditions guaranteeing the Input-to-State stability (ISS) and Incremental Input-to-State stability (dISS) of LSTM are derived. These properties are then exploited to design an observer with guaranteed convergence of the state estimate to the true one. Such observer is then embedded in a MPC scheme solving the tracking problem. The resulting closed-loop scheme is proved to be asymptotically stable. The training algorithm and control scheme are tested numerically on the simulator of a pH reactor, and the reported results confirm the effectiveness of the proposed approach.
Subjects: Systems and Control (eess.SY)
Cite as: arXiv:1910.04024 [eess.SY]
  (or arXiv:1910.04024v3 [eess.SY] for this version)

Submission history

From: Enrico Terzi [view email]
[v1] Wed, 9 Oct 2019 14:35:47 GMT (256kb,D)
[v2] Tue, 29 Oct 2019 11:01:05 GMT (257kb,D)
[v3] Thu, 27 Aug 2020 18:34:56 GMT (309kb,D)

Link back to: arXiv, form interface, contact.